International Business is a term used to collectively

describe all commercial transactions (private and

governmental, sales, investments, logistics, and
transportation) that take place between two or more nations.
Usually, private companies undertake such transactions for
profit; governments undertake them for profit and for
political reasons. It refers to all those business activities
which involves cross border transactions of goods, services,
resources between two or more nations. Transaction of
economic resources include capital, skills, people etc. for
international production of physical goods and services such
as finance, banking, insurance, construction etc.

A multinational enterprise (MNE) is a company that has a

worldwide approach to markets and production or one with
operations in more than a country. An MNE is often called
multinational corporation (MNC) or transnational company
(TNC). Well known MNCs include fast food companies such
as McDonald's and Yum Brands, vehicle manufacturers such
as General Motors, Ford Motor Company and Toyota,
consumer electronics companies like Samsung, LG and
Sony, and energy companies such as ExxonMobil, Shell and
BP. Most of the largest corporations operate in multiple
national markets.

Areas of study within this topic include differences in legal

systems, political systems, economic policy, language,
accounting standards, labor standards, living standards,
environmental standards, local culture, corporate culture,
foreign exchange market, tariffs, import and export
regulations, trade agreements, climate, education and many
more topics. Each of these factors requires significant
changes in how individual business units operate from one
country to the next.

The conduct of international operations depends on

companies' objectives and the means with which they carry
them out. The operations affect and are affected by the
physical and societal factors and the competitive
environment.

Operations

* Objectives: sales expansion, resource acquisition, risk

minimization

Means

* Modes: importing and exporting, tourism and

transportation, licensing and franchising, turnkey operations,
management contracts, direct investment and portfolio
investments.
* Functions: marketing, global manufacturing and supply
chain management, accounting, finance, human resources
 * Overlaying alternatives: choice of countries,
organization and control mechanisms

Physical and societal factors

* Political policies and legal practices

* Cultural factors
* Economic forces
* Geographical influences

Competitive factors

* Major advantage in price, marketing, innovation, or

other factors.
* Number and comparative capabilities of competitors
* Competitive differences by country

There has been growth in globalization in recent decades due

to the following eight factors:

* Technology is expanding, especially in transportation

and communications.
* Governments are removing international business
restrictions.
* Institutions provide services to ease the conduct of
international business.
* Consumers know about and want foreign goods and
services.
* Competition has become more global.
* Political relationships have improved among some major
economic powers.
* Countries cooperate more on transnational issues.
* Cross-national cooperation and agreements.

Studying international business is important because:

 * Most companies are either international or compete with
international companies.
* Modes of operation may differ from those used
domestically.
* The best way of conducting business may differ by
country.
* An understanding helps you make better career
decisions.
* An understanding helps you decide what governmental
policies to support.

Managers in international business must understand social

science disciplines and how they affect all functional
business fields.

Tom Travis, the managing partner of Sandler, Travis &

Rosenberg, PA. and international trade and customs
consultant, uses the Six Tenets when giving advice on how to
globalize one's business. The Six Tenets are as follows[3]:

1. Take advantage of trade agreements: think outside the

border
* Familiarize yourself with preference programs and
trade agreements.
* Read the fine print.
* Participate in the process.
* Seize opportunities when they arise.
2. Protect your brand at all costs
* You and your brand are inseparable.
* You must be vigilant in protecting your intellectual
property both at home and abroad.
* You must be vigilant in enforcing your IP rights.
* Protect your worldwide reputation by strict
adherence to labor and human rights standards.
3. Maintain high ethical standards
* Strong ethics translate into good business.
 * Forge ethical strategic partnerships.
* Understand corporate accountability laws.
* Become involved with the international business
self-regulation movement.
* Develop compliance protocols for import and export
operations.
* Memorialize your company's code of ethics and
compliance practices in writing.
* Appoint a leader.
4. Stay secure in an insecure world
* Security requires transparency throughout the supply
chain.
* Participate in trade-government partnerships.
* Make the most of new security measures.
* Secure your data.
* Keep your personnel secure.
5. Expect the Unexpected
* The unexpected will happen.
 * Do your research now.
* Address your particular circumstances.
6. All global business is personal
* Go to the source.
* Keep communications open.
* Keep the home office operational.
* Fly the flag at your overseas locations.
* Relate to offshore associates on a personal level.
* Be available to overseas clients and customers 24/7.
International Business in India looks really lucrative and
every passing day, it is coming up with only more
possibilities. The growth in the international business sector
in India is more than 7% annually. There is scope for more
improvement if only the relations with the neighboring
countries are stabilized. The mind-blowing performance of
the stock market in India has gathered all the more attention
(in comparison to the other international bourses). India
definitely stands as an opportune place to explore business
possibilities, with its high-skilled manpower and budding
middle class segment.

With the diverse cultural setup, it is advisable not to

formulate a uniform business strategy in India. Different
parts of the country are well-known for its different traits.
The eastern part of India is known as the ‘Land of the
intellectuals’, whereas the southern part is known for its
‘technology acumen’. On the other hand, the western part is
known as the ‘commercial-capital of the country’, with the
northern part being the ‘hub of political power’. With such
diversities in all the four segments of the country,
international business opportunity in India is surely huge.

Sectors having potential for International business in India –

1. Information Technology and Electronics Hardware.

2. Telecommunication.
 3. Pharmaceuticals and Biotechnology.
4. R&D.
5. Banking, Financial Institutions and Insurance &
Pensions.
6. Capital Market.
7. Chemicals and Hydrocarbons.
8. Infrastructure.
9. Agriculture and Food Processing.
10. Retailing.
11. Logistics.
12. Manufacturing.
13. Power and Non-conventional Energy.

Sectors like Health, Education, Housing, Resource

Conservation & Management Group, Water Resources,
Environment, Rural Development, Small and Medium
Enterprises (SME) and Urban Development are still not
tapped properly and thus the huge scope should be exploited.
 To foster the international business scenario in India,
bodies like CII, FICCI and the various Chambers of
Commerce, have a host of services like –
1. These bodies work closely with the Government and
the different business promotion organizations to infuse more
business development in India.
2. They help to build strong relationships with the
different international business organizations and the
multinational corporations.
3. These bodies help to identify the bilateral business
co-operation potential and thereafter make apt policy
recommendations to the different overseas Governments.
4. With opportunities huge, the International Business
trend in India is mind boggling. India International Business
community along with the domestic business community is
striving towards a steady path to be the Knowledge Capital
of the world.
 It was evident till a few years back that India had a
marginal role in the international affairs. The image was not
bright enough to be the cynosure among the shining stars.
The credit rating agencies had radically brought down the
country's ratings. But, as of now, after liberalization process
and the concept of an open economy – international business
in India grew manifold. Future definitely has more to offer to
the entire world.

Global Depository Receipts(GDR) / American Deposit

Receipts (ADR) / Foreign Currency Convertible Bonds
(FCCB)

Foreign Investment through ADRs/GDRs, Foreign Currency

Convertible Bonds (FCCBs) is treated as Foreign Direct
Investment. Indian companies are allowed to raise equity
capital in the international market through the issue of
GDR/ADRs/FCCBs. These are not subject to any ceilings on
investment. An applicant company seeking Government's
approval in this regard should have a consistent track record
for good performance (financial or otherwise) for a minimum
period of 3 years. This condition can be relaxed for
infrastructure projects such as power generation,
telecommunication, petroleum exploration and refining,
ports, airports and roads.

There is no restriction on the number of GDRs/ADRs/FCCBs

to be floated by a company or a group of companies in a
financial year. There is no such restriction because a
company engaged in the manufacture of items covered under
Automatic Route is likely to exceed the percentage limits
under Automatic Route, whose direct foreign investment
after a proposed GDRs/ADRs/FCCBs is likely to exceed 50
per cent/51 per cent/74 per cent as the case may be.
There are no end-use restrictions on GDRs/ADRs issue
proceeds, except for an express ban on investment in real
estate and stock markets. The FCCB issue proceeds need to
conform to external commercial borrowing end use
requirements. In addition, 25 per cent of the FCCB proceeds
can be used for general corporate restructuring.

ADR, GDR norms further relaxed

* Indian bidders allowed to raise funds through ADRs,

GDRs and external commercial borrowings (ECBs) for
acquiring shares of PSEs in the first stage and buying shares
from the market during the open offer in the second stage.
* Conversion and reconversion (a.k.a. two-way conversion
or fungibility) of shares of Indian companies into depository
receipts listed in foreign bourses, while extending tax
incentives to non-resident investors, allowed. The re-
coversion of ADRs/GDRs would, however, be governed by
the Foreign Exchange Management Act notified by the
Reserve Bank of India in March 2001.
* Permission to retain ADR/GDR proceeds abroad for
future foreign exchange requirements, removal of the
existing limit of $20,000 for remittance under the employees
stock option scheme (ESOP) and permitting remittance up to
$ 1 million from proceeds of sales of assets here.
* Companies have been allowed to invest 100 per cent of
the proceeds of ADR/GDR issues (as against the earlier
ceiling of 50%) for acquisitions of foreign companies and
direct investments in joint ventures and wholly-owned
subsidiaries overseas.
* Any Indian company which has issued ADRs/GDRs
may acquire shares of foreign companies engaged in the
same area of core activity upto $100 million or an amount
equivalent to ten times of their exports in a year, whichever
is higher. Earlier, this facility was available only to Indian
companies in certain sectors.
* FIIs can invest in a company under the portfolio
investment route upto 24 per cent of the paid-up capital of
the company. It can be increased to 40% with approval of
general body of the shareholders by a special resolution. This
limit has now been increased to 49% from the present 40%.
* Two way fungibility in ADR/GDR issues of Indian
companies has been introduced subject to sectoral caps
wherever applicable. Stock brokers in India can now
purchase shares and deposit these with the Indian custodian
for issue of ADRs/GDRs by the overseas depository to the
extent of the ADRs/GDRs that have been converted into
underlying shares.

On Fungibility

Two-way fungibility of ADRs/GDRs issued by Indian

Companies was permitted by the Government of India and
the RBI. The RBI has now, vide APDIR Circular No: 21
dated February 13th 2002, issued operative guidelines for the
2 way fungibility of ADR / GDR.

Earlier, once a company issued ADR / GDR, and if the

holder wanted to obtain the underlying equity shares of the
Indian Company, then, such ADR / GDR would be converted
into shares of the Indian Company. Once such conversion
took place, it was not possible to reconvert the equity shares
into ADR / GDR. The present rules of the RBI make such
reconversion possible, to the extent of ADR / GDR which
have been converted into equity shares and sold in the local
market. This would take place in the following manner:

# Stock Brokers in India have been authorized to purchase

shares of Indian Companies for reconversion
# The Domestic Custodian would coordinate with the
Overseas Depository and the Indian Company to verify the
quantum of reconversion which is possible and also to ensure
that the sectoral cap is not breached.
# The Domestic Custodian would then inform the Overseas
Depository to issue ADR / GDR to the overseas Investor.

Re-issue of ADRs/GDRs would be permitted to the extent of

ADRs/GDRs that have been redeemed and the underlying
shares sold in the domestic market. Two-way fungibility
implies that an investor who holds ADRs/GDRs can cancel
them with the depository and sell the underlying shares in the
market. The company can then issue fresh ADRs to the
extent of shares cancelled.

No specific permission of the RBI will be required for the re-

conversion. Besides, investments under foreign currency
convertible bonds and ordinary shares will be treated as
direct foreign investment. Accordingly, the re-conversion of
shares into ADRs/GDRs will be distinct from portfolio
investments by foreign institutional investors (FIIs). The RBI
guidelines state that the transactions will be demand-driven
and would not require company involvement or fresh
permissions. The custodian would monitor the re-issuance of
ADRs/GDRs within the sectoral cap fixed by the
Government. Each purchase transaction will be only against
delivery and payment received in foreign exchange through
banking channels. For this purpose, all SEBI registered
brokers will be able to act as intermediaries in the two-way
fungibility of ADRs/GDRs.

The RBI has already given general permission to brokers

(not banks, but SEBI registered stockbrokers. RBI has
conveyed general permission through a Notification
No.FEMA.41/2001-RB dated 2nd March 2001, for these
brokers to buy shares on behalf of the overseas investor) to
buy shares on behalf of overseas investors (this include both
foreign investors as well as domestic shareholders). As
secondary market operations, the acquisition of shares on
behalf of the overseas investors through the intermediary
would fall within the regulatory purview of Securities and
Exchange Board of India (SEBI). The Central bank has said
that since the demand for re-conversion of shares into
ADRs/GDRs would be from overseas investors and not the
company, the expenses would be borne by the investor. The
transactions will be governed by the Income-Tax Act.

Benefits of fungibility
The key benefits that could accrue to investors (ADR/GDR
holders and domestic investors) and companies from two-
way fungibility are: improvement in liquidity and elimination
of arbitrage.

The conventional definition of liquidity is the ease with

which an asset (in this case, ADRs/GDRs) can be bought or
sold quickly with relatively small price changes. This
essentially means that a liquid market for a security must
have depth and breadth, and aid speedy price discovery. A
liquid market is said to have depth if buy and sell orders exist
both above and below the prices (at which a stock or
ADR/GDR) is transacting. Similarly, the market is said to
have breadth if buy and sell orders exist in good volume.

In the one-way fungible regime, ADRs/GDRs suffered from

price volatility and liquidity problems, basically for two
reasons. The first reason was the low ADR issue size that
accounted for low free-float in the US market and, thereby,
low trading volumes in the security.

Second, the GDR market had been largely dormant (with the
exception of a few high-profile stocks) for the past couple of
years. This affected the depth, breadth and price-discovery
process of GDRs in these markets. Two-way fungibility may
at least revive some market interest in these stocks.

Reduction/elimination of arbitrage
In an efficient market, two assets with identical attributes
must sell for the same price, and so should an identical asset
trading in two different markets. If the prices of such an asset
differ, a profitable opportunity arises to sell the asset where it
is overpriced and buy it back where it is under priced.
Obviously, arbitrageurs (speculators aiming to exploit these
riskless opportunities) can step in and exploit this profit
opportunity.

Under the one-way fungibility regime, though identical

assets (namely stocks in the domestic market and
ADRs/GDRs in the overseas markets) traded at different
prices (at a discount/premium), the arbitrage opportunities
went a begging because of restrictions on the capital account.
By introducing two-way fungibility, market forces may
trigger a realignment of prices, minimising the widely
divergent premium/discount levels prevailing between
ADR/GDR prices and the domestic stock prices.

Euro Issues by Indian Companies

Earlier, Indian Companies required approval of the

Government of India before issue of Foreign Currency
Convertible Bonds (FCCBs). The RBI, has vide FEMA
Notification No : 55 dated March 7th 2002, liberalised these
rules. Accordingly:
# Indian Companies seeking to raise FCCBs are permitted to
raise them under the Automatic Route upto US 50 Million
Dollars per financial year without any approval.
# The FCCBs raised shall be subject to the sectoral limits*
prescribed by the Government of India.
# Maturity period for the FCCBs shall be at least 5 years and
the "all in cost" at least 100 basis points less than that
prescribed for External Commercial Borrowings.

Some restrictions had been imposed previously on the

number of issues that could be floated by an individual
company or a group of companies during a financial year.
There will henceforth be no restrictions on the number of
Euro-Issues to be floated by a company or a group of
companies in a financial year.

GDR end-uses will include:

# financing capital goods imports;
# capital expenditure including domestic
purchase/installation of plant, equipment and buildings and
investments in software development;
# prepayment or scheduled repayment of earlier external
borrowings;
# investments abroad where these have been approved by
competent authorities;
# equity investment in JVs/WOSs in India. However,
investments in stock markets and real estate will not be
permitted. Up to a maximum of 25 per cent of the total
proceeds may be used for general corporate restructuring,
including working capital requirements of the company
raising the GDR.

Currently, companies are permitted to access foreign capital

market through Foreign Currency Convertible Bonds for
restructuring of external debt that helps to lengthen maturity
and soften terms, and for end-use of funds which conform to
the norms prescribed for the Government for External
Commercial Borrowings (ECB) from time to time. In
addition to these, not more than 25 per cent of FCCB issue
proceeds may be used for general corporate restructuring
including working capital requirements.

FCCBs are available and accessible more freely as compared

to external debt, and the expectation of the Government is
that FCCBs should have a substantially finer spread than
ECBs. Accordingly, the all-in costs for FCCBs should be
significantly better than the corresponding debt instruments
(ECBs). Companies will not be permitted to issue warrants
along with their Euro-issue. The policy and guidelines for
Euro-issues will be subject to review periodically.

Sectoral Caps

As per the Foreign Investment guidelines issued by the

Government of India, Ministry of Industry, foreign
investment (equity/preference shares) upto certain specified
limits would be permitted by Reserve Bank under Automatic
Route as under:
In relaxation of earlier guidelines, GDR end - uses will
include :
# Foreign investment (equity/preference) upto 50% in respect
of Mining activities;
# Foreign investment (equity/preference) upto 51% in (i)
industries/items included in part 'B' of Annexure III to
Ministry of Industry's Press Note No.14 (1997 series) dated
8th October 1997** and (ii) a trading company primarily
engaged in export activity; in software development
# Foreign investment (equity/preference) upto 74% in
industries/items included in part 'C' of Annexure III to
Ministry of Industry's Press Note No.14 (1997 series) dated
8th October 1997**
# Foreign Investment upto 100% in industries/items included
in Part 'D' of Annexure III, to Ministry of Industry's Press
Note No.14 (1997 Series)** as amended from time to time
provided the foreign investment in a project does not exceed
Rs.1500 crores.

Global Depositary Receipt - GDR

What Does It Mean?
What Does Global Depositary Receipt - GDR Mean?
1. A bank certificate issued in more than one country for
shares in a foreign company. The shares are held by a foreign
branch of an international bank. The shares trade as domestic
shares, but are offered for sale globally through the various
bank branches.

2. A financial instrument used by private markets to raise

capital denominated in either U.S. dollars or euros.
Investopedia Says
Investopedia explains Global Depositary Receipt - GDR
1. A GDR is very similar to an American Depositary Receipt.

2. These instruments are called EDRs when private markets

are attempting to obtain euros.
Foreign Investment through GDRs

Global Depository Receipt (GDRs) enables the elevation of

equity capital of Indian companies in the global market.
GDRs are usually calculated in foreign dollars and not
subjected to any other currency or ceilings in terms of
investments. The Indian companies who are seeking
government's approval for enrolling in this sector are
required to have a good track record in terms of financial
performance or other matters for a minimum period of three
years. For infrastructure projects such telecommunication,
power generation, petroleum exploration and refining,
airports, roads, and ports, this requirement of three years is
not mandatory.
Foreign Investment through GDRs - Clearance from FIPB-
FIPB is responsible for the clearance of Euro-issue to be
floated by a company or a group of companies in the
financial year. For example, a manufacturing company,
which is entitled to the regulations and norms of Annex-III of
the New Industrial Policy, is likely to witness a 51 percent
increase in the foreign direct investment amount after the
proposed Euro issue. The manufacturing company therefore
is required to seek the FIPB approval before obtaining the
final approval from Ministry of Finance in case of a project,
which does not abide by Annex-III of the New Industrial
Policy.
Use of GDRs in Foreign Direct Investment-
The proceeds of GDRs are used for providing financial
assistance to import capital goods, capital expenditure of
domestic purchase or installation of plants, instrument and
constructing and investment in software development
procedures, defrayal or scheduled repayment of earlier
external loans, and equity investment in JV/WOSs in India.
Restrictions in Foreign Investment through GDRs-
Foreign Direct Investments in stock markets and real estate
market through GDRs are not allowed. The proceeds of
investments can be stopped either in the foreign country from
where the FDI is coming or in India depending on the use of
funds that are being approved for end uses. FDI investments
under every circumstance require prior approval of the Indian
government.

Recent Change in the Dimensions of India's Foreign

Trade

Whole world has recognized India as super power of 21st

century. India is youngest county in the world growing a rate
of more than 8 percent. Large population of India provides
market to the countries of the world. At the same time it
provides opportunities to India in terms of extracting the
potentials of its manpower and other resources to emerge as
real super power. India's foreign trade should also reflect her
potentials to emerge as a super power.

Economy of India
The economy of India is the twelfth largest economy in the
world by nominal value[9] and the fourth largest by
purchasing power parity (PPP).[10] In the 1990s, following
economic reform from the socialist-inspired economy of
post-independence India, the country began to experience
rapid economic growth, as markets opened for international
competition and investment. In the 21st century, India is an
emerging economic power with vast human and natural
resources, and a huge knowledge base. Economists predict
that by 2020,[11] India will be among the leading economies
of the world.

India was under social democratic-based policies from 1947

to 1991. The economy was characterised by extensive
regulation, protectionism, and public ownership, leading to
pervasive corruption and slow growth.[12][13][14][15] Since
1991, continuing economic liberalisation has moved the
economy towards a market-based system.[13][14] A revival
of economic reforms and better economic policy in 2000s
accelerated India's economic growth rate. By 2008, India had
established itself as the world's second-fastest growing major
economy.[16][17][18] However, the year 2009 saw a
significant slowdown in India's official GDP growth rate to
6.1%[19] as well as the return of a large projected fiscal
deficit of 10.3% of GDP which would be among the highest
in the world.[20][21]

India's large service industry accounts for 62.6% of the

country's GDP while the industrial and agricultural sector
contribute 20% and 17.5% respectively. Agriculture is the
predominant occupation in India, accounting for about 52%
of employment. The service sector makes up a further 34%,
and industrial sector around 14%.[22] The labor force totals
half a billion workers. Major agricultural products include
rice, wheat, oilseed, cotton, jute, tea, sugarcane, potatoes,
cattle, water buffalo, sheep, goats, poultry and fish.[23]
Major industries include telecommunications, textiles,
chemicals, food processing, steel, transportation equipment,
cement, mining, petroleum, machinery, information
technology enabled services and software.[23]

India's per capita income (nominal) is $1032, ranked 139th in

the world,[24] while its per capita (PPP) of US$2,932 is
ranked 128th.[25][26] Previously a closed economy, India's
trade has grown fast.[13] India currently accounts for 1.5%
of World trade as of 2007 according to the WTO. According
to the World Trade Statistics of the WTO in 2006, India's
total merchandise trade (counting exports and imports) was
valued at $294 billion in 2006 and India's services trade
inclusive of export and import was $143 billion. Thus, India's
global economic engagement in 2006 covering both
merchandise and services trade was of the order of $437
billion, up by a record 72% from a level of $253 billion in
2004. India's trade has reached a still relatively moderate
share 24% of GDP in 2006, up from 6% in 1985.[13]

Despite robust economic growth, India continues to face

many major problems. The recent economic development has
widened the economic inequality across the country.[27]
Despite sustained high economic growth rate, approximately
80% of its population lives on less than $2 a day (PPP).[28]
Even though the arrival of Green Revolution brought end to
famines in India,[29] 40% of children under the age of three
are underweight and a third of all men and women suffer
from chronic energy deficiency.[30]

History
India's economic history can be broadly divided into three
eras, beginning with the pre-colonial period lasting up to the
18th century. The advent of British colonisation started the
colonial period in the early 19th century, which ended with
independence in 1947. The third period stretches from
independence in 1947 until now.

Pre-colonial

The citizens of the Indus Valley civilisation, a permanent

settlement that flourished between 2800 BC and 1800 BC,
practiced agriculture, domesticated animals, used uniform
weights and measures, made tools and weapons, and traded
with other cities. Evidence of well planned streets, a drainage
system and water supply reveals their knowledge of urban
planning, which included the world's first urban sanitation
systems and the existence of a form of municipal
government.[33]

The 1872 census revealed that 99.3% of the population of the

region constituting present-day India resided in villages,[34]
whose economies were largely isolated and self-sustaining,
with agriculture the predominant occupation. This satisfied
the food requirements of the village and provided raw
materials for hand-based industries, such as textiles, food
processing and crafts. Although many kingdoms and rulers
issued coins, barter was prevalent. Villages paid a portion of
their agricultural produce as revenue to the rulers, while its
craftsmen received a part of the crops at harvest time for
their services.[35]

Religion, especially Hinduism, and the caste and the joint

family systems, played an influential role in shaping
economic activities.[36] The caste system functioned much
like medieval European guilds, ensuring the division of
labour, providing for the training of apprentices and, in some
cases, allowing manufacturers to achieve narrow
specialization. For instance, in certain regions, producing
each variety of cloth was the specialty of a particular sub-
caste.
Estimates of the per capita income of India (1857–1900) as
per 1948–49 prices.[37]

Textiles such as muslin, Calicos, shawls, and agricultural

products such as pepper, cinnamon, opium and indigo were
exported to Europe, the Middle East and South East Asia in
return for gold and silver.[38]

Assessment of India's pre-colonial economy is mostly

qualitative, owing to the lack of quantitative information.
One estimate puts the revenue of Akbar's Mughal Empire in
1600 at £17.5 million, in contrast with the total revenue of
Great Britain in 1800, which totalled £16 million.[39] India,
by the time of the arrival of the British, was a largely
traditional agrarian economy with a dominant subsistence
sector dependent on primitive technology. It existed
alongside a competitively developed network of commerce,
manufacturing and credit. After the decline of the Mughals,
western, central and parts of south and north India were
integrated and administered by the Maratha Empire. The
Maratha Empire's budget in 1740s, at its peak, was Rs. 100
million. After the loss at Panipat, the Maratha Empire
disintegrated into confederate states of Gwalior, Baroda,
Indore, Jhansi, Nagpur, Pune and Kolhapur. Gwalior state
had a budget of Rs. 30M. However, at this time, British East
India company entered the Indian political theatre. Until
1857, when India was firmly under the British crown, the
country remained in a state of political instability due to
internecine wars and conflicts.[4

Colonial
An aerial view of Calcutta Port taken in 1945. Calcutta,
which was the economic hub of British India, saw increased
industrial activity during World War II.

Company rule in India brought a major change in the

taxation environment from revenue taxes to property taxes,
resulting in mass impoverishment and destitution of majority
of farmers and led to numerous famines.[41] The economic
policies of the British Raj effectively bankrupted India's large
handicrafts industry and caused a massive drain of India's
resources.[42][43] Indian Nationalists employed the
successful Swadeshi movement, as strategy to diminish
British economic superiority by boycotting British products
and the reviving the market for domestic-made products and
production techniques. India had become a strong market for
superior finished European goods. This was because of vast
gains made by the Industrial revolution in Europe, the effects
of which was deprived to Colonial India.
The Nationalists had hoped to revive the domestic industries
that were badly effected by polices implemented by British
Raj which had made them uncompetitive to British made
goods.

An estimate by Cambridge University historian Angus

Maddison reveals that "India's share of the world income fell
from 22.6% in 1700, comparable to Europe's share of 23.3%,
to a low of 3.8% in 1952".[44] It also created an institutional
environment that, on paper, guaranteed property rights
among the colonizers, encouraged free trade, and created a
single currency with fixed exchange rates, standardized
weights and measures, capital markets. It also established a
well developed system of railways and telegraphs, a civil
service that aimed to be free from political interference, a
common-law and an adversarial legal system.[45] India's
colonisation by the British coincided with major changes in
the world economy—industrialisation, and significant growth
in production and trade. However, at the end of colonial rule,
India inherited an economy that was one of the poorest in the
developing world,[46] with industrial development stalled,
agriculture unable to feed a rapidly growing population,
India had one of the world's lowest life expectancies, and low
rates for literacy.

The impact of the British rule on India's economy is a

controversial topic. Leaders of the Indian independence
movement, and left-nationalist economic historians have
blamed colonial rule for the dismal state of India's economy
in its aftermath and that financial strength required for
Industrial development in Europe was derived from the
wealth taken from Colonies in Asia and Africa. At the same
time right-wing historians have countered that India's low
economic performance was due to various sectors being in a
state of growth and decline due to changes brought in by
colonialism and a world that was moving towards
industrialization and economic integration

Independence to 1991

Compare India (orange) with South Korea (yellow). Both

started from about the same income level in 1950. The graph
shows GDP per capita of South Asian economies and South
Korea as a percent of the American GDP per capita.

Indian economic policy after independence was influenced

by the colonial experience (which was seen by Indian leaders
as exploitative in nature) and by those leaders' exposure to
Fabian socialism. Policy tended towards protectionism, with
a strong emphasis on import substitution, industrialization,
state intervention in labor and financial markets, a large
public sector, business regulation, and central planning.[48]
Five-Year Plans of India resembled central planning in the
Soviet Union. Steel, mining, machine tools, water,
telecommunications, insurance, and electrical plants, among
other industries, were effectively nationalized in the mid-
1950s.[49] Elaborate licences, regulations and the
accompanying red tape, commonly referred to as Licence
Raj, were required to set up business in India between 1947
and 1990.[50]

Jawaharlal Nehru, the first prime minister, along with the

statistician Prasanta Chandra Mahalanobis, carried on by
Indira Gandhi formulated and oversaw economic policy.
They expected favorable outcomes from this strategy,
because it involved both public and private sectors and was
based on direct and indirect state intervention, rather than the
more extreme Soviet-style central command system.[51]
[dead link] The policy of concentrating simultaneously on
capital- and technology-intensive heavy industry and
subsidizing manual, low-skill cottage industries was
criticized by economist Milton Friedman, who thought it
would waste capital and labour, and retard the development
of small manufacturers.[52][dead link] The rate from 1947–
80 was derisively referred to as the Hindu rate of growth,
because of the unfavourable comparison with growth rates in
other Asian countries, especially the "East Asian Tigers".[45]

The Rockefeller Foundation's research in high-yielding

varieties of seeds, their introduction after 1965 and the
increased use of fertilizers and irrigation are known
collectively as the Green Revolution in India, which
provided the increase in production needed to make India
self-sufficient in food grains, thus improving agriculture in
India. Famine in India, once accepted as inevitable, has not
returned since the end of colonial.

Since 1991
In the late 80s, the government led by Rajiv Gandhi eased
restrictions on capacity expansion for incumbents, removed
price controls and reduced corporate taxes. While this
increased the rate of growth, it also led to high fiscal deficits
and a worsening current account. The collapse of the Soviet
Union, which was India's major trading partner, and the first
Gulf War, which caused a spike in oil prices, caused a major
balance-of-payments crisis for India, which found itself
facing the prospect of defaulting on its loans.[54] India asked
for a $1.8 billion bailout loan from IMF, which in return
demanded reforms.[55]

In response, Prime Minister Narasimha Rao along with his

finance minister Dr. Manmohan Singh initiated the economic
liberalisation of 1991. The reforms did away with the
Licence Raj (investment, industrial and import licensing) and
ended many public monopolies, allowing automatic approval
of foreign direct investment in many sectors.[56] Since then,
the overall direction of liberalisation has remained the same,
irrespective of the ruling party, although no party has tried to
take on powerful lobbies such as the trade unions and
farmers, or contentious issues such as reforming labour laws
and reducing agricultural subsidies.[57] Since 1990 India has
emerged as one of the fastest-growing economies in the
developing world; during this period, the economy has
grown constantly, but with a few major setbacks. This has
been accompanied by increases in life expectancy, literacy
rates and food security.

While the credit rating of India was hit by its nuclear tests in
1998, it has been raised to investment level in 2007 by S&P
and Moody's.[58] In 2003, Goldman Sachs predicted that
India's GDP in current prices will overtake France and Italy
by 2020, Germany, UK and Russia by 2025 and Japan by
2035. By 2035, it was projected to be the third largest
economy of the world, behind US and China.[59][60] In
2009 India purchased 200 Tons of Gold for $6.7 Billion from
IMF as a total role reversal from 1991.
Sectors
Agriculture
Farmers work inside a rice field in Andhra Pradesh. India is
the second largest producer of rice in the world after
China[61] and Andhra Pradesh is the 2nd largest rice
producing state in India with West Bengal being the largest.
[62]
Main articles: Agriculture in India, Forestry in India, Animal
husbandry in India, and Fishing in India

India ranks second worldwide in farm output. Agriculture

and allied sectors like forestry, logging and fishing accounted
for 16.6% of the GDP in 2007, employed 60% of the total
workforce[22] and despite a steady decline of its share in the
GDP, is still the largest economic sector and plays a
significant role in the overall socio-economic development of
India. Yields per unit area of all crops have grown since
1950, due to the special emphasis placed on agriculture in the
five-year plans and steady improvements in irrigation,
technology, application of modern agricultural practices and
provision of agricultural credit and subsidies since Green
revolution in India. However, international comparisons
reveal the average yield in India is generally 30% to 50% of
the highest average yield in the world.[63]

India is the largest producer in the world of milk, cashew

nuts, coconuts, tea, ginger, turmeric and black pepper.[64] It
also has the world's largest cattle population: 193 million.
[65] It is the second largest producer of wheat, rice, sugar,
cotton, silk, peanuts and inland fish.[66] It is the third largest
producer of tobacco.[66] India is the largest fruit producer,
accounting for 10% of the world fruit production. It is the
leading producer of bananas, sapotas and mangoes.[66]
India is the second largest producer and the largest consumer
of silk in the world, with the majority of the 77 million kg
(2005)[67] production taking place in Karnataka State,
particularly in Mysore and the North Bangalore regions of
Muddenahalli, Kanivenarayanapura, and Doddaballapura, the
upcoming sites of a INR 700 million "Silk City".[68][69]

Industry and services

India has one of the world's fastest growing automobile

industries[72][73] Shown here is Tata Motors' Nano, world's
least expensive car in production.[74]

Industry accounts for 54.6% of the GDP and employ 17% of

the total workforce.[22] However, about one-third of the
industrial labour force is engaged in simple household
manufacturing only.[75][dead link] In absolute terms, India
is 16th in the world in terms of nominal factory output.[76]
India's small industry makes up 5% of carbon dioxide
emissions in the world.

Economic reforms brought foreign competition, led to

privatisation of certain public sector industries, opened up
sectors hitherto reserved for the public sector and led to an
expansion in the production of fast-moving consumer goods.
[77] Post-liberalisation, the Indian private sector, which was
usually run by oligopolies of old family firms and required
political connections to prosper was faced with foreign
competition, including the threat of cheaper Chinese imports.
It has since handled the change by squeezing costs,
revamping management, focusing on designing new products
and relying on low labour costs and technology.[78]

Textile manufacturing is the second largest source for

employment after agriculture and accounts for 26% of
manufacturing output.[79] Tirupur has gained universal
recognition as the leading source of hosiery, knitted
garments, casual wear and sportswear.[80] Dharavi slum in
Mumbai has gained fame for leather products. Tata Motors'
Nano attempts to be the world's cheapest car.[74]

India is fifteenth in services output. It provides employment

to 23% of work force, and it is growing fast, growth rate
7.5% in 1991–2000 up from 4.5% in 1951–80. It has the
largest share in the GDP, accounting for 55% in 2007 up
from 15% in 1950.[22]

Business services (information technology, information

technology enabled services, business process outsourcing)
are among the fastest growing sectors contributing to one
third of the total output of services in 2000. The growth in
the IT sector is attributed to increased specialization, and an
availability of a large pool of low cost, but highly skilled,
educated and fluent English-speaking workers, on the supply
side, matched on the demand side by an increased demand
from foreign consumers interested in India's service exports,
or those looking to outsource their operations. The share of
India's IT industry to the country's GDP increased from 4.8
% in 2005-06 to 7% in 2008.[81][82] In 2009, seven Indian
firms were listed among the top 15 technology outsourcing
companies in the world.[83] In March 2009, annual revenues
from outsourcing operations in India amounted to US$60
billion and this is expected to increase to US$225 billion by
2020.[84]

Organized retail such supermarkets accounts for 24% of the

market as of 2008.[85] Regulations prevent most foreign
investment in retailing. Moreover, over thirty regulations
such as "signboard licences" and "anti-hoarding measures"
may have to be complied before a store can open doors.
There are taxes for moving goods to states, from states, and
even within states.[85]
Tourism in India is relatively undeveloped, but growing at
double digits. Some hospitals woo medical tourism.[86]

Banking and finance

The Indian money market is classified into: the organised

sector (comprising private, public and foreign owned
commercial banks and cooperative banks, together known as
scheduled banks); and the unorganised sector (comprising
individual or family owned indigenous bankers or money
lenders and non-banking financial companies (NBFCs)). The
unorganised sector and microcredit are still preferred over
traditional banks in rural and sub-urban areas, especially for
non-productive purposes, like ceremonies and short duration
loans.[88]
Mumbai is the financial and commercial capital of India.
Shown here is the World Trade Center of Mumbai

Prime Minister Indira Gandhi nationalised 14 banks in 1969,

followed by six others in 1980, and made it mandatory for
banks to provide 40% of their net credit to priority sectors
like agriculture, small-scale industry, retail trade, small
businesses, etc. to ensure that the banks fulfill their social
and developmental goals. Since then, the number of bank
branches has increased from 10,120 in 1969 to 98,910 in
2003 and the population covered by a branch decreased from
63,800 to 15,000 during the same period. The total deposits
increased 32.6 times between 1971 to 1991 compared to 7
times between 1951 to 1971. Despite an increase of rural
branches, from 1,860 or 22% of the total number of branches
in 1969 to 32,270 or 48%, only 32,270 out of 5 lakh
(500,000) villages are covered by a scheduled bank.[89][90]
The public sector banks hold over 75% of total assets of the
banking industry, with the private and foreign banks holding
18.2% and 6.5% respectively.[91] Since liberalisation, the
government has approved significant banking reforms. While
some of these relate to nationalised banks (like encouraging
mergers, reducing government interference and increasing
profitability and competitiveness), other reforms have
opened up the banking and insurance sectors to private and
foreign players.[22][92]

More than half of personal savings are invested in physical

assets such as land, houses, cattle, and gold.[93]

Natural resources

India has the world's fifth largest wind power industry, with
an installed wind power capacity of 9,587 MW. Shown here
is a wind farm in Muppandal, Tamil Nadu.
India's total cultivable area is 1,269,219 km² (56.78% of total
land area), which is decreasing due to constant pressure from
an ever growing population and increased urbanisation. India
has a total water surface area of 314,400 km² and receives an
average annual rainfall of 1,100 mm. Irrigation accounts for
92% of the water utilisation, and comprised 380 km² in 1974,
and is expected to rise to 1,050 km² by 2025, with the
balance accounted for by industrial and domestic consumers.
India's inland water resources comprising rivers, canals,
ponds and lakes and marine resources comprising the east
and west coasts of the Indian ocean and other gulfs and bays
provide employment to nearly 6 million people in the
fisheries sector. In 2008, India had the world's third largest
fishing industry.[94]

India's major mineral resources include coal, iron,

manganese, mica, bauxite, titanium, chromite, limestone and
thorium. India meets most of its domestic energy demand
through its 92 billion tonnes of coal reserves (about 10% of
world's coal reserves).[95]

India's huge thorium reserves — about 25% of world's

reserves — is expected to fuel the country's ambitious
nuclear energy program in the long-run. India's dwindling
uranium reserves stagnated the growth of nuclear energy in
the country for many years.[96] However, the Indo-US
nuclear deal has paved the way for India to import uranium
from other countries.[97] India is also believed to be rich in
certain renewable sources of energy with significant future
potential such as solar, wind and biofuels (jatropha,
sugarcane).
[edit] Petroleum and Natural gas
ONGC platform at Bombay High in the Arabian Sea. As of
2010, India is the world's fifth largest consumer of oil.[98]
India's oil reserves, found in Bombay High, parts of Gujarat,
Rajasthan and eastern Assam, meet 25% of the country's
domestic oil demand.[22][99] India's total proven oil reserves
stand at 11 billion barrels,[100] of which Bombay High is
believed to hold 6.1 billion barrels[101] and Mangala Area in
Rajasthan an additional 3.6 billion barrels.[102]

In 2009, India imported 2.56 million barrels of oil per day,

making it one of largest buyers of crude oil in the world.
[103] The petroleum industry in India mostly consists of
public sector companies such as Oil and Natural Gas
Corporation (ONGC), Hindustan Petroleum Corporation
Limited (HPCL) and Indian Petrochemicals Corporation
Limited (IPCL). There are some major private Indian
companies in oil sector such as Reliance Industries Limited
(RIL) which operates the world's largest oil refining
complex.[104]
Pharmaceuticals

India has a self reliant Pharmaceuticals industry. The

majority of its medical consumables are produced
domestically. Pharmaceutical Industry in India is dotted with
companies like Ranbaxy Laboratories, Dr. Reddy's
Laboratories, Cipla which have created a niche for
themselves at world level.

Today, India is an exporter of countries like United States

and Russia. In terms of the global market, India currently
holds a modest 1-2% share, but it has been growing at
approximately 10% per year. Indian Pharmaceutical Industry
is often compared to Pharmaceutical Industry in the USA.

External trade and investment

India's economy is mostly dependent on its large internal
market with external trade accounting for just 20% of the
country's GDP.[106] In 2008, India accounted for 1.45% of
global merchandise trade and 2.8% of global commercial
services export.[107] Until the liberalization of 1991, India
was largely and intentionally isolated from the world
markets, to protect its economy and to achieve self-reliance.
Foreign trade was subject to import tariffs, export taxes and
quantitative restrictions, while foreign direct investment
(FDI) was restricted by upper-limit equity participation,
restrictions on technology transfer, export obligations and
government approvals; these approvals were needed for
nearly 60% of new FDI in the industrial sector. The
restrictions ensured that FDI averaged only around US$200
million annually between 1985 and 1991; a large percentage
of the capital flows consisted of foreign aid, commercial
borrowing and deposits of non-resident Indians.[108] India's
exports were stagnant for the first 15 years after
independence, due to the predominance of tea, jute and
cotton manufactures, demand for which was generally
inelastic. Imports in the same period consisted predominantly
of machinery, equipment and raw materials, due to nascent
industrialization.

Since liberalization, the value of India's international trade

has become more broad-based and has risen to Rs.
63,080,109 crores in 2003–04 from Rs.1,250 crores in 1950–
51. India's major trading partners are China, the US, the
UAE, the UK, Japan and the EU.[109] The exports during
April 2007 were $12.31 billion up by 16% and import were
$17.68 billion with an increase of 18.06% over the previous
year.[110] In 2006-07, major export commodities included
engineering goods, petroleum products, chemicals and
pharmaceuticals, gems and jewellery, textiles and garments,
agricultural products, iron ore and other minerals. Major
import commodities included crude oil and related products,
machinery, electronic goods, gold and silver.[111]

India is a founding-member of General Agreement on Tariffs

and Trade (GATT) since 1947 and its successor, the WTO.
While participating actively in its general council meetings,
India has been crucial in voicing the concerns of the
developing world. For instance, India has continued its
opposition to the inclusion of such matters as labour and
environment issues and other non-tariff barriers into the
WTO policies.[112]
[edit] Balance of payments
Cumulative Current Account Balance 1980-2008 based on
the IMF data

Since independence, India's balance of payments on its

current account has been negative. Since liberalisation in the
1990s (precipitated by a balance of payment crisis), India's
exports have been consistently rising, covering 80.3% of its
imports in 2002–03, up from 66.2% in 1990–91. India's
growing oil import bill is seen as the main driver behind the
large current account deficit.[113] In 2007-08, India
imported 120.1 million tonnes of crude oil, more than 3/4th
of the domestic demand, at a cost of $61.72 billion.[114]

Although India is still a net importer, since 1996–97 its

overall balance of payments (i.e., including the capital
account balance) has been positive, largely on account of
increased foreign direct investment and deposits from non-
resident Indians; until this time, the overall balance was only
occasionally positive on account of external assistance and
commercial borrowings. As a result, India's foreign currency
reserves stood at $285 billion in 2008.

Due to the global late-2000s recession, both Indian exports

and imports declined by 29.2% and 39.2% respectively in
June 2009.[115] The steep decline was because countries hit
hardest by the global recession, such as United States and
members of the European Union, account for more than 60%
of Indian exports.[116] However, since the decline in imports
was much sharper compared to the decline in exports, India's
trade deficit reduced to 252.5 billion rupee.[115]

India's reliance on external assistance and commercial

borrowings has decreased since 1991–92, and since 2002–03,
it has gradually been repaying these debts. Declining interest
rates and reduced borrowings decreased India's debt service
ratio to 4.5% in 2007.[117] In India, External Commercial
Borrowings (ECBs) are being permitted by the Government
for providing an additional source of funds to Indian
corporates. The Ministry of Finance monitors and regulates
these borrowings (ECBs) through ECB policy guidelines.
[118]
[edit] Foreign direct investment in India
Share of top five investing countries in FDI inflows. (2000–
2007)[119] Rank Country Inflows
(Million USD) Inflows (%)
1 Mauritius 85,178 44.24%[120]
2 United States 18,040 9.37%
3 United Kingdom 15,363 7.98%
4 Netherlands 11,177 5.81%
5 Singapore 9,742 5.06%

As the fourth-largest economy in the world in PPP terms,

India is a preferred destination for foreign direct investments
(FDI);[121] India has strengths in telecommunication,
information technology and other significant areas such as
auto components, chemicals, apparels, pharmaceuticals, and
jewellery. Despite a surge in foreign investments, rigid FDI
policies resulted in a significant hindrance. However, due to
some positive economic reforms aimed at deregulating the
economy and stimulating foreign investment, India has
positioned itself as one of the front-runners of the rapidly
growing Asia Pacific Region.[121] India has a large pool of
skilled managerial and technical expertise. The size of the
middle-class population stands at 300 million and represents
a growing consumer market.[122]

The inordinately high investment from Mauritius is due to

routing of international funds through the country given
significant capital gains tax advantages; double taxation is
avoided due to a tax treaty between India and Mauritius, and
Mauriitus is a capital gains tax haven, effectively creating a
zero-taxation FDI channel.

India's recently liberalized FDI policy (2005) allows up to a

100% FDI stake in ventures. Industrial policy reforms have
substantially reduced industrial licensing requirements,
removed restrictions on expansion and facilitated easy access
to foreign technology and foreign direct investment FDI. The
upward moving growth curve of the real-estate sector owes
some credit to a booming economy and liberalized FDI
regime. In March 2005, the government amended the rules to
allow 100 per cent FDI in the construction business.[123]
This automatic route has been permitted in townships,
housing, built-up infrastructure and construction
development projects including housing, commercial
premises, hotels, resorts, hospitals, educational institutions,
recreational facilities, and city- and regional-level
infrastructure.

A number of changes were approved on the FDI policy to

remove the caps in most sectors. Fields which require
relaxation in FDI restrictions include civil aviation,
construction development, industrial parks, petroleum and
natural gas, commodity exchanges, credit-information
services and mining. But this still leaves an unfinished
agenda of permitting greater foreign investment in politically
sensitive areas such as insurance and retailing. FDI inflows
into India reached a record $19.5 billion in fiscal year 2006-
07 (April-March), according to the government's Secretariat
for Industrial Assistance. This was more than double the total
of US$7.8bn in the previous fiscal year. The FDI inflow for
2007-08 has been reported as $24 billion[124] and for 2008-
09, it is expected to be above $35 billion.[125] A critical
factor in determining India's continued economic growth and
realizing the potential to be an economic superpower is going
to depend on how the government can create incentives for
FDI flow across a large number of sectors in India.[126]

Currency

The Indian rupee is the only legal tender accepted in India.

The exchange rate as on 23 March 2010 is 45.40 INR the
USD,[127] 61.45 to a EUR, and 68.19 to a GBP. The Indian
rupee is accepted as legal tender in the neighboring Nepal
and Bhutan, both of which peg their currency to that of the
Indian rupee. The rupee is divided into 100 paise. The
highest-denomination banknote is the 1,000 rupee note; the
lowest-denomination coin in circulation is the 25 paise coin
(it earlier had 1, 2, 5, 10 and 20 paise coins which have been
discontinued by the Reserve Bank of India).[128]

The Rupee hit a record low during early 2009 on account of

global recession. However, due to a strong domestic market,
India managed to bounce back sooner than the western
countries. Since September 2009 there has been a constant
appreciation in Rupee versus most Tier 1 currencies. On 11
January 2010 Rupee went as high as 45.50 to a United states
dollar and on 10 January 2010 as high as Rupee 73.93 to a
British Pound. A rising rupee also prompted Government of
India to buy 200 tonnes of Gold from IMF.
The RBI, the country's central bank was established on 1
April 1935. It serves as the nation's monetary authority,
regulator and supervisor of the financial system, manager of
exchange control and as an issuer of currency. The RBI is
governed by a central board, headed by a governor who is
appointed by the Central government of India.

Income and consumption

* 85.7% of the population lives on less than $2.50 (PPP) a

day, down from 92.5% in 1981. This is much higher than the
80.5% in Sub-Saharan Africa.[129]
* 75.6% of the population lives on less than $2 a day
(PPP), which is around 20 rupees or $0.5 a day in nominal
terms. It was down from 86.6%, but is still even more than
the 73.0% in Sub-Saharan Africa.[129][130][131][132][133]
 * 24.3% of the population earned less than $1 (PPP,
around $0.25 in nominal terms) a day in 2005, down from
42.1% in 1981.[129][134]
* 41.6% of its population is living below the new
international poverty line of $1.25 (PPP) per day, down from
59.8% in 1981.[129] The World Bank further estimates that a
third of the global poor now reside in India.

Today, more people can afford a bicycle than ever before.

Some 40% of Indian households owns a bicycle, with
ownership rates ranging from around 30% to 70% at state
level.[135] Housing is modest. According to Times of India,
"a majority of Indians have per capita space equivalent to or
less than a 10 feet x 10 feet room for their living, sleeping,
cooking, washing and toilet needs." and "one in every three
urban Indians lives in homes too cramped to exceed even the
minimum requirements of a prison cell in the US."[136] The
average is 103 sq ft (9.6 m2) per person in rural areas and
117 sq ft (10.9 m2) per person in urban areas.[136]

Around half of Indian children are malnourished. The

proportion of underweight children is nearly double that of
Sub-Saharan Africa.[30][137] However, India has not had
famines since the Green Revolution in the early 1970s. While
poverty in India has reduced significantly, official figures
estimate that 27.5%[138] of Indians still lived below the
national poverty line of $1 (PPP, around 10 rupees in
nominal terms) a day in 2004-2005.[139] A 2007 report by
the state-run National Commission for Enterprises in the
Unorganised Sector (NCEUS) found that 65% of Indians, or
750 million people, lived on less than 20 rupees per day[140]
with most working in "informal labour sector with no job or
social security, living in abject poverty."[141]

Since the early 1950s, successive governments have

implemented various schemes, under planning, to alleviate
poverty, that have met with partial success. All these
programmes have relied upon the strategies of the Food for
work programme and National Rural Employment
Programme of the 1980s, which attempted to use the
unemployed to generate productive assets and build rural
infrastructure.[142] In August 2005, the Indian parliament
passed the Rural Employment Guarantee Bill, the largest
programme of this type in terms of cost and coverage, which
promises 100 days of minimum wage employment to every
rural household in all the India's 600 districts. The question
of whether economic reforms have reduced poverty or not
has fuelled debates without generating any clear cut answers
and has also put political pressure on further economic
reforms, especially those involving the downsizing of labour
and cutting agricultural subsidies.[143][144]
[edit] Employment
See also: Indian labour law
Agricultural and allied sectors accounted for about 60% of
the total workforce in 2003 same as in 1993–94. While
agriculture has faced stagnation in growth, services have
seen a steady growth. Of the total workforce, 8% is in the
organised sector, two-thirds of which are in the public sector.
The NSSO survey estimated that in 1999–2000, 106 million,
nearly 10% of the population were unemployed and the
overall unemployment rate was 7.3%, with rural areas doing
marginally better (7.2%) than urban areas (7.7%). India's
labor force is growing by 2.5% annually, but employment
only at 2.3% a year.[145]

Official unemployment exceeds 9%. Regulation and other

obstacles have discouraged the emergence of formal
businesses and jobs. Almost 30% of workers are casual
workers who work only when they are able to get jobs and
remain unpaid for the rest of the time.[145] Only 10% of the
workforce is in regular employment.[145] India's labor
regulations are heavy even by developing country standards
and analysts have urged the government to abolish them.[13]
[146]

Unemployment in India is characterized by chronic or

disguised unemployment. Government schemes that target
eradication of both poverty and unemployment (which in
recent decades has sent millions of poor and unskilled people
into urban areas in search of livelihoods) attempt to solve the
problem, by providing financial assistance for setting up
businesses, skill honing, setting up public sector enterprises,
reservations in governments, etc. The decreased role of the
public sector after liberalization has further underlined the
need for focusing on better education and has also put
political pressure on further reforms.[142][147]

Child labor is a complex problem that is basically rooted in

poverty. The Indian government is implementing the world's
largest child labor elimination program, with primary
education targeted for ~250 million. Numerous non-
governmental and voluntary organizations are also involved.
Special investigation cells have been set up in states to
enforce existing laws banning employment of children
(under 14) in hazardous industries. The allocation of the
Government of India for the eradication of child labor was
$10 million in 1995-96 and $16 million in 1996-97. The
allocation for 2007 is $21 million.[148]

In 2006, remittances from Indian migrants overseas made up

$27 billion or about 3% of India's GDP.[149]

Economic trends

India's 300 million strong middle-class population is growing

at an annual rate of 5%.[150] Shown here is a residential area
in the Mumbai metropolitan area.
In the revised 2007 figures, based on increased and
sustaining growth, more inflows into foreign direct
investment, Goldman Sachs predicts that "from 2007 to
2020, India’s GDP per capita in US$ terms will quadruple",
and that the Indian economy will surpass the United States
(in US$) by 2043.[15] In spite of the high growth rate, the
report stated that India would continue to remain a low-
income country for decades to come but could be a "motor
for the world economy" if it fulfills its growth potential.[15]
Goldman Sachs has outlined 10 things that it needs to do in
order to achieve its potential and grow 40 times by 2050.
These are

1. improve governance
2. raise educational achievement
3. increase quality and quantity of universities
4. control inflation
 5. introduce a credible fiscal policy
6. liberalize financial markets
7. increase trade with neighbours
8. increase agricultural productivity
9. improve infrastructure and
10. improve environmental quality.[151]

Issues
Agriculture
An Indian farmer
Main article: Agriculture in India

Slow agricultural growth is a concern for policymakers as

some two-thirds of India’s people depend on rural
employment for a living. Current agricultural practices are
neither economically nor environmentally sustainable and
India's yields for many agricultural commodities are low.
Poorly maintained irrigation systems and almost universal
lack of good extension services are among the factors
responsible. Farmers' access to markets is hampered by poor
roads, rudimentary market infrastructure, and excessive
regulation.

– World Bank: "India Country Overview 2008"[152]

The low productivity in India is a result of the following

factors:

* According to "India: Priorities for Agriculture and Rural

Development" by World Bank, India's large agricultural
subsidies are hampering productivity-enhancing investment.
Overregulation of agriculture has increased costs, price risks
and uncertainty. Government interventions in labor, land, and
credit markets are hurting the market. Infrastructure and
services are inadequate.[153]
* Illiteracy, slow progress in implementing land reforms
and inadequate or inefficient finance and marketing services
for farm produce.
* The average size of land holdings is very small (less than
20,000 m²) and is subject to fragmentation, due to land
ceiling acts and in some cases, family disputes. Such small
holdings are often over-manned, resulting in disguised
unemployment and low productivity of labour.
* Adoption of modern agricultural practices and use of
technology is inadequate, hampered by ignorance of such
practices, high costs and impracticality in the case of small
land holdings.
* World Bank says that the allocation of water is
inefficient, unsustainable and inequitable. The irrigation
infrastructure is deteriorating.[153] Irrigation facilities are
inadequate, as revealed by the fact that only 52.6% of the
land was irrigated in 2003–04,[154] which result in farmers
still being dependent on rainfall, specifically the Monsoon
season. A good monsoon results in a robust growth for the
economy as a whole, while a poor monsoon leads to a
sluggish growth.[155] Farm credit is regulated by NABARD,
which is the statutory apex agent for rural development in the
subcontinent.

India has many farm insurance companies that insure wheat,

fruit, rice and rubber farmers in the event of natural disasters
or catastrophic crop failure, under the supervision of the
Ministry of Agriculture. One notable company that provides
all of these insurance policies is Agriculture Insurance
Company of India and it alone insures almost 20 million
farmers.

India's population is growing faster than its ability to produce

rice and wheat.[29] The most important structural reform for
self-sufficiency is the ITC Limited plan to connect 20,000
villages to the Internet by 2013.[156] This will provide
farmers with up to date crop prices for the first time, which
should minimise losses incurred from neighbouring
producers selling early and in turn facilitate investment in
rural areas.
[edit] Corruption
Overview of the index of perception of corruption, 2007
Main article: Corruption in India

Corruption has been one of the pervasive problems affecting

India. The economic reforms of 1991 reduced the red tape,
bureaucracy and the Licence Raj that had strangled private
enterprise and was blamed by Chakravarthi Rajagopalachari
for the corruption and inefficiencies. Yet, a 2005 study by
Transparency International (TI) India found that more than
half of those surveyed had firsthand experience of paying
bribe or peddling influence to get a job done in a public
office.[157]

The Right to Information Act (2005) and equivalent acts in

the Indian states, that require government officials to furnish
information requested by citizens or face punitive action,
computerisation of services and various central and state
government acts that established vigilance commissions have
considerably reduced corruption or at least have opened up
avenues to redress grievances.[157] The 2009 report by
Transparency International ranks India at 84th place and
states that significant improvements were made by India in
reducing corruption.[158][159]
[edit] Government
Main article: Government of India
See also: Taxation in India and Corruption in India
The number of people employed in non-agricultural
occupations in the public and private sectors. Totals are
rounded. Private sector data relates to non-agriculture
establishments with 10 or more employees.[142]

The current government has concluded that most spending

fails to reach its intended recipients.[160] Lant Pritchett calls
India's public sector "one of the world's top ten biggest
problems — of the order of AIDS and climate change".[160]
The Economist's 2008 article about the Indian civil service
stated that the Indian central government employs around 3
million people, including "vast armies of paper-shuffling
peons".[160]

At local level, administration can be worse. It is not unheard

of that a majority of a state's assembly seats can be held by
convicted criminals.[161] One study found that 25% of
public sector teachers and 40% of public sector medical
workers could not be found at the workplace. India's absence
rates are one of the worst in the world.[162][163][164][165]

The government of India is highly indebted and its former

investment-grade status has deteriorated to near junk status.
[166] India's current public-debt to GDP ratio is 60% (US is
53%).[167]
[edit] Education
Main article: Education in India

India has made huge progress in terms of increasing primary

education attendance rate and expanding literacy to
approximately two thirds of the population.[168] The right to
education at elementary level has been made one of the
fundamental rights under the Eighty-Sixth Amendment of
2002.[169] However, education is still far behind developing
countries such as China and continues to face challenges.
Despite growing investment in education, 40% of the
population is illiterate and only 15% of the students reach
high school.[170] An optimistic estimate is that only one in
five job-seekers in India has ever had any sort of vocational
training.[171]
[edit] Infrastructure
See also: Transport in India, Indian Road Network, Ports in
India, Electricity in India, States of India by installed power
capacity, Water supply and sanitation in India, and
Communications in India
Rapid increases in exports has resulted in congestion on
highways across India. Shown here is the Yashwantrao
Chavan expressway in Maharashtra.

Development of infrastructure was completely in the hands

of the public sector and was plagued by corruption,
bureaucratic inefficiencies, urban-bias and an inability to
scale investment.[172] India's low spending on power,
construction, transportation, telecommunications and real
estate, at $31 billion or 6% of GDP in 2002 had prevented
India from sustaining higher growth rates. This has prompted
the government to partially open up infrastructure to the
private sector allowing foreign investment[142][173][174]
which has helped in a sustained growth rate of close to 9%
for the past six quarters.[175]
Some 600 million Indians have no mains electricity at all.
[176] While 80% of Indian villages have at least an
electricity line, just 44% of rural households have access to
electricity.[177] According to a sample of 97,882 households
in 2002, electricity was the main source of lighting for 53%
of rural households compared to 36% in 1993.[178] Some
half of the electricity is stolen, compared with 3% in China.
The stolen electricity amounts to 1.5% of GDP.[177][179]
Almost all of the electricity in India is produced by the
public sector. Power outages are common.[176] Many buy
their own power generators to ensure electricity supply. As
of 2005 the electricity production was at 661.6 billion kWh
with oil production standing at 785,000 bbl/day. In 2007,
electricity demand exceeded supply by 15%.[176] Multi
Commodity Exchange has tried to get a permit to offer
electricity future markets.[180]
Indian Road Network is developing. Trucking goods from
Gurgaon to the port in Mumbai can take up to 10 days.[181]
India has the world's third largest road network.[182]
Container traffic is growing at 15% a year.[183] Some 60%
of India’s container traffic is handled by the Jawaharlal
Nehru Port Trust in Mumbai. Internet use is rare; there were
only 7.57 million broadband lines in India in November
2009, however it is still growing at slower rate and is
expected to boom after the launch of 3G and wimax services.
[184]

Most urban cities have good water supply water 24 hours a

day, while some smaller cities face water shortages in
summer season. A World Bank report says it is an
institutional problem in water agencies, or "how the agency
is embedded in the relationships between politics and the
citizens who are the consumers."[185]
[edit] Labour laws
Main article: Indian labour laws

India’s labor regulations — among the most restrictive and

complex in the world — have constrained the growth of the
formal manufacturing sector where these laws have their
widest application. Better designed labor regulations can
attract more labor- intensive investment and create jobs for
India’s unemployed millions and those trapped in poor
quality jobs. Given the country’s momentum of growth, the
window of opportunity must not be lost for improving the job
prospects for the 80 million new entrants who are expected to
join the work force over the next decade.

– World Bank: India Country Overview 2008.[152]

India's restrictive labor regulations hamper the large-scale

creation of formal industrial jobs.[13][171][186] A recent
report highlights a growing labor unrest all over India which
is hampering the industrial output.[187]

India ranked 133th on the Ease of Doing Business Index

2010, behind countries such as China (89th), Pakistan (85th),
and Nigeria (125th). The Constitution provides protection of
child labour, slavory, equality of opportunities and forced
labor etc. in form of fundamental rights, but the
implementation of provisions cited is a big question mark.
[188]
[edit] Economic disparities
Main articles: Economic disparities in India and Poverty in
India

Lagging states need to bring more jobs to their people by

creating an attractive investment destination. Reforming
cumbersome regulatory procedures, improving rural
connectivity, establishing law and order, creating a stable
platform for natural resource investment that balances
business interests with social concerns, and providing rural
finance are important.

– World Bank: India Country Overview 2008[152]

Slums next to high-rise commercial buildings in Kaloor,

Kochi. Hundreds of people, mostly comprising migrant
labourers who come to the city seeking job prospects, reside
in such shabby areas.[189]

One of the critical problems facing India's economy is the

sharp and growing regional variations among India's
different states and territories in terms of per capita income,
poverty, availability of infrastructure and socio-economic
development.[190] Seven low-income states - Bihar,
Chhattisgarh, Jharkhand, Madhya Pradesh, Orissa, Rajasthan,
and Uttar Pradesh - are home to more than half of India's
population.[191]
Between 1999 and 2008, the annualized growth rates for
Maharashtra (9%),[192] Gujarat (8.8%), Haryana (8.7%), or
Delhi (7.4%) were much higher than for Bihar (5.1%), Uttar
Pradesh (4.4%), or Madhya Pradesh (3.5%).[193]

Poverty rates in rural Orissa (43%) and rural Bihar (40%) are
some of the worst in the world.[185] On the other hand, rural
Haryana (5.7%) and rural Punjab (2.4%) compare well with
middle-income countries.[185]

The five-year plans have attempted to reduce regional

disparities by encouraging industrial development in the
interior regions, but industries still tend to concentrate
around urban areas and port cities[194] After liberalization,
the more advanced states are better placed to benefit from
them, with infrastructure like well developed ports,
urbanisation and an educated and skilled workforce which
attract manufacturing and service sectors. The union and
state governments of backward regions are trying to reduce
the disparities by offering tax holidays, cheap land, etc., and
focusing more on sectors like tourism, which although being
geographically and historically determined, can become a
source of growth and is faster to develop than other sectors.
[195][196]
[edit] Environment and health
Calicut Medical College, Kerala
Main articles: Environment of India, Water supply and
sanitation in India, and Healthcare in India

On Yale and Columbia's Environmental Performance Index,

India's score is 21/100 on sanitation, compared with 67/100
for the region and 48/100 for the country income group.[197]
The Export-Import Bank of India, also known as Exim Bank
of India, is the leading export finance institution in the
country. The bank was set up in the year 1982 under the
Export-Import Bank of India Act 1981. The Government of
India launched the Export-Import Bank Of India with an aim
to augment exports from India and also to combine the
country's foreign trade and investment with the overall
economic growth. The bank began its operations as a
supplier of export credit, but has over the period evolved into
an institution that plays a major role in partnering Indian
Industries including small and medium enterprises.

Export-Import Bank Of India Initiatives

Export-Import Bank of India has been one of the prime

institutions that encourages project exports from India. The
bank offers wide-ranging services for enhancing the prospect
of Indian project exports. Exim Bank's Overseas Investment
Finance program gives a variety of facilities for Indian
reserves and acquirements overseas. The facilities consist of
direct equity participation by the bank in the overseas
venture and non-funded activities by the overseas venture
and loan to the Indian companies for equity participation in
the venture abroad. As part of Exim Bank's marketing
Finance Program, the bank offers support to small and
medium enterprises in their export marketing efforts
consisting of financing the soft expenditure linking to
completion of tactical and systematic export market
development plans.
Export-Import Bank Of India Objective

The primary objective of the Export-Import Bank of India is

to provide financial assistance to importers and exporters and
function as the top financial institution. Some of the services
of the bank include: overseas investment finance, film
finance, export credit, finance for export oriented units and
agricultural & SME finance. In the period of 2005- 2006 the
total amount of loan given out by the bank amounted to
150,389 million, while this figure shot up to Rs. 220,760
million in the flowing year.
Export-Import Bank Of India and Foreign Trade

Export-Import Bank of India plays the role of source of

finance, promoter, coordinator and consultation to India's
Foreign Trade. The bank is the coordinator of the Working
Group Mechanism for the clearance of projects, service
exports and deferred payment exports. This group comprises
of Exim Bank and Government of India representatives from
the Ministries of Finance, Commerce and external Affairs,
Export Credit Guarantee Corporation of India Ltd,
commercial banks that are certified foreign exchange dealers
and the Reserve Bank of India. This working group gives
clearance to contracts sponsored by Exim Bank or
commercial banks and operates as a single window
mechanism for clearance of export proposal terms
www.speedbit.comhttp://max.fileflyer.comwww.fileflyer.co
mhttp://test.fileflyer.comhttp://localhost/FileFlyersearch.spee
dbit.comstart.speedbit.comhttp://search.speedbit.comhttp://st
art.speedbit.com
PDATE name="update" xmlformatversion="1"/>
www.speedbit.comhttp://max.fileflyer.comwww.fileflye
r.comhttp://test.fileflyer.comhttp://localhost/FileFlyersearch.s
peedbit.comstart.speedbit.comhttp://search.speedbit.comhttp:
//start.speedbit.com

Organization

Exim Bank is managed by a Board of Directors, which has

representatives from the Government, Reserve Bank of India,
Export Credit Guarantee Corporation (ECGC) of India, a
financial institution, public sector banks, and the business
community.

The Bank's functions are segmented into several operating

groups including:

Corporate Banking Group which handles a variety of

financing programmes for Export Oriented Units (EOUs),
Importers, and overseas investment by Indian companies.

Project Finance / Trade Finance Group handles the entire

range of export credit services such as supplier's credit, pre-
shipment Agri Business Group, to spearhead the initiative to
promote and support Agri-exports. The Group handles
projects and export transactions in the agricultural sector for
financing.

Small and Medium Enterprise: The group handles credit

proposals from SMEs under various lending programmes of
the Bank.

Export Services Group offers variety of advisory and value-

added information services aimed at investment promotion

Export Marketing Services Bank offers assistance to Indian

companies, to enable them establish their products in
overseas markets.

Besides these, the Support Services groups, which include:

Research & Planning, Corporate Finance, Loan Recovery,
Internal Audit, Management Information Services,
Information Technology, Legal, Human Resources
Management and Corporate Affairs.
[edit] External linksRole of EXIM Bank of India in Foreign
Investment
This article focuses on the Role of EXIM Bank of India in
Foreign Investment. EXIM Bank is a fully owned bank by
the Government of India and is governed by the Board of
Directors and gets tax return from the government, various
financial institutions, banks and business communities.
An Overview of EXIM Bank-
The main objective of Export-Import Bank (EXIM Bank) is
to provide financial assistance to promote the export
production in India. The financial assistance provided by the
EXIM Bank widely includes the following:

* Direct financial assistance

* Foreign investment finance
* Term loaning options for export production and export
development
* Pre-shipping credit
* Buyer's credit
* Lines of credit
 * Reloaning facility
* Export bills rediscounting
* Refinance to commercial banks

The Export-Import Bank also provides non-funded facility in

the form of guarantees to the Indian exporters.
Various Stages of Exports Covered by EXIM Bank-

* Development of export makers

* Expansion of export production capacity
* Production for exports
* Financing post-shipment activities
* Export of manufactured goods
* Export of projects
* Export of technology and softwares

Forms of Financial Assistance Provided by EXIM Bank

to Indian Companies-
 * Delayed Payment Exports- Term loans are provided to
those exporters who deal with exporting of goods and
services and this enables them to offer delayed credit to the
foreign buyers. This system of deferred credit covers Indian
consultancies, technology, and other services. Commercial
banks take part in this program either directly or under risk
syndication arrangements.
* Pre-shipment credit- Indian companies which are highly
involved in the execution of export activities beyond the
cycle time of six months are funded by EXIM Bank. The
construction or turnkey project exporters enjoy the provision
of rupee mobilization.
* Term loans for export production- EXIM Bank offers
term loans to the 100 percent export oriented units, units
involved in free trade zones, and exporters of various
softwares in India. EXIM bank also works in association
with International Finance Corporation, Washington, to
provide financial assistance to the small scale and medium
industrial units in terms of ameliorating the export
production capacity of these units in India. EXIM Bank also
provides funded and non- funded facilities to deemed exports
from India.
* Foreign Investment Finance- EXIM bank provides
financial assistance for equity contribution to the Indian
companies who form Joint Venture with the foreign
companies.
* Financing export marketing- It helps the exporters carry
out their export market development plan in Indian market.

Financial Assistance Provided by EXIM Bank to Overseas

Companies-

* Foreign Buyer's Credit- the foreign players are entitled

to a sum of financial assistance in order to import goods and
services on deferred payments
 * Lines of Credit- EXIM bank also offers financial
assistance to the overseas financial institutions and various
government agencies for import of goods and services from
India.
* Reloaning Options to Foreign Banks- The foreign banks
are entrusted with funding from EXIM bank in order to
provide the same to the their clients across the globe for
importing of goods from India

Imports and exports are the two important components of a

foreign trade. Foreign trade is the exchange of goods and
services between the two countries, across their international
borders.'Imports' imply the physical movement of goods into
a country from another country in a legal manner. It refers to
the goods that are produced abroad by foreign producers and
are used in the domestic economy to cater to the needs of the
domestic consumers. Similarly, 'exports' imply the physical
movement of goods out of a country in a legal manner. It
refers to the goods that are produced domestically in a
country and are used to cater to the needs of the consumers in
foreign countries. Thus, the imports and exports have made
the world a local market. The country which is purchasing
the goods is known as the importing country and the country
which is selling the goods is known as the exporting country.
The traders involved in such transactions are importers and
exporters respectively.

In India, exports and imports are regulated by the Foreign

Trade (Development and Regulation) Act, 1992, which
replaced the Imports and Exports(Control) Act, 1947, and
gave the Government of India enormous powers to control it.
The salient features of the Act are as follows:-

* It has empowered the Central Government to make

provisions for development and regulation of foreign trade
by facilitating imports into, and augmenting exports from
India and for all matters connected therewith or incidental
thereto.

* The Central Government can prohibit, restrict and

regulate exports and imports, in all or specified cases as well
as subject them to exemptions.

* It authorizes the Central Government to formulate and

announce an Export and Import (EXIM) Policy and also
amend the same from time to time, by notification in the
Official Gazette.

* It provides for the appointment of a Director General of

Foreign Trade by the Central Government for the purpose of
the Act. He shall advise Central Government in formulating
export and import policy and implementing the policy.

* Under the Act, every importer and exporter must obtain

a 'Importer Exporter Code Number' (IEC) from Director
General of Foreign Trade or from the officer so authorised.

* The Director General or any other officer so authorised

can suspend or cancel a licence issued for export or import of
goods in accordance with the Act. But he does it after giving
the licence holder a reasonable opportunity of being heard.

* As per the provisions of the Act , the Government of

India formulates and announces an Export and Import policy
(EXIM policy) and amends it from time to time. EXIM
policy refers to the policy measures adopted by a country
with reference to its exports and imports. Such a policy
become particularly important in a country like India, where
the import and export of items plays a crucial role not just in
balancing budgetary targets, but also in the over all economic
development of the country. The principal objectives of the
policy are:-

o To facilitate sustained growth in exports of the

country so as to achieve larger percentage share in the global
merchandise trade.

o To provide domestic consumers with good quality

goods and services at internationally competitive prices as
well as creating a level playing field for the domestic
producers.
 o To stimulate sustained economic growth by
providing access to essential raw materials, intermediates,
components, consumables and capital goods required for
augmenting production and providing services.

o To enhance the technological strength and efficiency

of Indian agriculture, industry and services, thereby
improving their competitiveness to meet the requirements of
the global markets.

o To generate new employment opportunities and to

encourage the attainment of internationally accepted
standards of quality.

Besides this Act, there are some other laws which control the
export and import of goods. These include:-

* Tea Act,1953

* Coffee Act, 1942

* The Rubber Act, 1947

* The Marine Products Export Development Authority

Act, 1972

* The Enemy Property Act, 1968

 * The Export (Quality Control and Inspection) Act, 1963

* The Tobacco Board Act, 1975

At the central level, the Ministry of Commerce and Industry

is the most important organ concerned with the promotion
and regulation of the foreign trade in India. The Ministry has
an elaborate organizational set up to look after the various
aspects of trade. Within the Ministry,the Department of
Commerce is responsible for formulating and implementing
the foreign trade policy. The Department is also entrusted
with responsibilities relating to multilateral and bilateral
commercial relations, state trading, export promotion
measures and development and regulation of certain export
oriented industries and commodities.The matters relating to
foreign trade are dealt with by the following divisions of the
Department :-
 1. Administrative and General Division
2. Finance Division
3. Economic Division
4. Trade Policy Division
5. Foreign Trade Territorial Division
6. Export Products Division
7. Export Industries Division
8. Export Services Division
9. Supply Division

The Department's jurisdiction extends over:-

(a) Two Attached Offices:-

* Directorate General of Foreign Trade (DGFT):- with its

headquarters at New Delhi, is headed by the Director General
of Foreign Trade. It is responsible for implementing the
Foreign Trade Policy/Exim Policy with the main objective of
promoting Indian exports. The DGFT also issues licences to
exporters and monitors their corresponding obligations
through a network of regional offices. The regional offices
are located at 33 places.

* Directorate General of Supplies and Disposal

(DGS&D):- with its headquarters at New Delhi, is headed by
the Director General. It functions as the executive arm of the
Supply Division of the Department of Commerce for
conclusion of Rate Contracts for common user items,
procurement of stores, inspection of stores, shipment and
clearance of imported stores/cargo. It has three Regional
Offices located at Chennai, Mumbai and Kolkata.

(b) Five Subordinate Offices:-

 * Directorate General of Commercial Intelligence and
Statistics (DGCI&S):- with its office located at Kolkata, is
headed by the Director General. It is entrusted with the work
of collecting, compiling and publishing/ disseminating trade
statistics and various types of commercial information
required by the policy makers, researchers, importers,
exporters, traders as well as overseas buyers.

* Office of Development Commissioner of Special

Economic Zones:- The Special Economic Zones (SEZs) are
geographically exclusive enclaves separated from domestic
tariff areas. The main objective of SEZs is to provide certain
common facilities and a duty free environment for exporters.
Each Zone is headed by a Development Commissioner and is
administered as per the SEZ scheme announced on 31st
March, 2000.
 * Office of the Custodian of Enemy Property (CEP):- is
located in Mumbai with a Branch office at Kolkata. The
office is functioning under the Enemy Property Act,1968. All
immovable (like land, buildings, etc.) and movable properties
(like securities, shares, debentures, bank balances, viz. fixed
deposits and other amounts lying in the enemy nationals'
bank accounts, Provident fund balances etc.) all over India
belonging to or held by or managed on behalf of Pakistani
nationals between the period 10.9.1965 and 26.9.1977 are
vested in the Custodian of Enemy Property for India.

* Pay and Accounts Office (Supply):- The payment and

accounting functions of Supply Division, including those of
DGS&D, are performed by the Chief Controller of Accounts
(CCA) under the Departmentalized Accounting System.
Payment to suppliers across the country is made through this
organisation.

* Pay and Accounts Office (Commerce & Textiles):- The

Pay and Accounts Office, common to both the Department of
Commerce and the Ministry of Textiles, is responsible for the
payment of claims, accounting of transactions and other
related matters through the four Departmental Pay &
Accounts Offices in Delhi, two in Mumbai, two in Kolkata
and one in Chennai.

(c) Ten Autonomous Bodies:-

* Coffee Board :- The Coffee Board of India is an

autonomous body, functioning under the Ministry of
Commerce and Industry, Government of India. The Board
serves as a guide of the coffee industry in India. The Board
focuses on research, development, extension, quality
upgradation, market information, and the domestic and
external promotion of Indian coffee.

* Rubber Board :- The board is engaged in the

development of the rubber industry. This is done by assisting
and encouraging scientific ,technical and economic research;
supplying technical advice to rubber growers; and training
growers in improved methods of plantation and cultivation.

* Tea Board :- The primary functions of tea board include

rendering financial and technical assistance for
cultivation,manufacture,marketing of tea; promoting tea
exports ;aiding research and developmental activities for
augmentation of tea production and improvement of tea
quality as well as encouraging and assisting small growers
sector financially and technically.
 * Tobacco Board:- The Government of India established
the Tobacco Board, in place of Tobacco Export Promotion
Council, under the Tobacco Board Act of 1975 to regulate
production, promotion of overseas marketing and to control
recurring instances of imbalances in supply and demand,
which lead to market problems,The Tobacco Board Act aims
at the planned development of Tobacco Industry in the
country. The activities of the Board includes the regulation of
the production and curing of Virginia Tobacco with regard to
the demand in India and abroad.

* Spices Board :- Spices Board was constituted on 26th

February 1986 under the Spices Board Act 1986. It is one of
the Commodity Boards functioning under the Ministry of
Commerce & Industry. It is an autonomous body responsible
for the export promotion of the scheduled spices and
production or development of some of them such as
Cardamom and Vanilla.

* Export Inspection Council (EIC), New Delhi :- The

Export Inspection Council is responsible for the enforcement
of quality control and compulsory preshipment inspection of
various commodities meant for export and notified under the
Export (Quality Control & Inspection) Act, 1963.

* Indian Institute of Foreign Trade (IIFT), NewDelhi :- is

engaged in the following activities:-

o Training of Personnel in modern techniques of

international trade;
 o Organisation of Research in problems of foreign
trade;

o Organisation of marketing research, area surveys,

commodity surveys, market surveys;

o Dissemination of information arising from its

activities relating to research and market studies.

* Indian Institute of Packaging (IIP), Mumbai :- is

registered under the Societies Registration Act.The main aim
of this Institute is to undertake research of raw materials for
the packaging industry, to organise training programmes on
packaging technology and to stimulate consciousness of the
need for good packaging etc.

* Marine Products Exports Development Authority

(MPEDA), Kochi :- functions under the Ministry of
Commerce, Government of India and acts as a coordinating
agency with different Central and State Government
establishments engaged in fishery production and allied
activities. The Authority is responsible for development of
the marine products industry with special focus on marine
exports. The role envisaged for the MPEDA is
comprehensive covering fisheries of all kinds, increasing
exports, specifying standards, processing, marketing,
extension and training in various aspects of the marine
industry.
 * Agricultural and Processed Food Products Export
Development Authority (APEDA), New Delhi :- came into
existence in 1986 to further develop agricultural
commodities and processed foods, and to promote their
exports.The aim is to maximize foreign exchange earnings
through increased agro exports, to provide better income to
the farmers through higher unit value realization and to
create employment opportunities in rural areas by
encouraging value added exports of farm produce.

(d) Export Promotion Councils (EPCs):-

Presently there are twelve EPCs under the administrative

control of the Ministry of Commerce.These councils are
registered as non-profit organisations under the Companies
Act.The Councils perform both the advisory and executive
functions.These councils are also the registering authorities
under the Import Policy for Registered Exporters.
(e) Other Organisations:-

* Federation of Indian Export Organisations (FIEO):- is an

apex body of various export promotion organizations and
institutions with its major regional offices at Delhi, Mumbai,
Chennai and Kolkata. It provides the content, direction and
thrust to India's global export effort.

* Indian Council of Arbitration (ICA), New Delhi :- set up

under the Societies Registration Act promotes arbitration as a
means of settling commercial disputes and popularises the
concepts of arbitration among the traders, particularly those
engaged in international trade.

* Indian Diamond Institute (IDI), Surat :- With the

objective of enhancing the quality, design and global
competitiveness of the Indian Jewellery, the Indian Diamond
Institute (IDI) was established as a pivotal institute for
imparting technical skills to the Gems and Jewellery industry
in the areas of Gemology and Jewellery manufacture.

(f) Advisory Bodies

* Board of Trade (BOT):- was set up on May 5, 1989 with

a view to providing an effective mechanism to maintain
continuous dialogue with trade and industry in respect of
major developments in the field of International Trade.

* Export Promotion Board (EPB):- provide policy and

infrastructural support through greater coordination amongst
concerned Ministries for boosting the growth of exports.
 * Directorate General of Anti-Dumping & Allied Duties
(DGAD):- The Directorate is responsible for carrying out
investigations and to recommend, where required, under
Customs Tariff Act, the amount of anti-dumping
duty/countervailing duty on the identified articles which
would be adequate to remove injury to the domestic industry.

(g) Public Sector Undertakings:-

The following trading/service corporations are functioning

under the administrative control of the Department of
Commerce:-

* State Trading Corporation (STC) of India Ltd.

* MMTC (Minerals and Metals Trading Corporation of

India) Limited,

* PEC Ltd,

* Export Credit Guarantee Corporation (ECGC) of India

Ltd.

* India Trade Promotion Organisation (ITPO)

^ Top
Foreign Trade Policy
In India, the main legislation concerning foreign trade is the
Foreign Trade (Development and Regulation) Act, 1992. The
Act provides for the development and regulation of foreign
trade by facilitating imports into, and augmenting exports
from, India and for matters connected therewith or incidental
thereto. As per the provisions of the Act, the Government :-
(i) may make provisions for facilitating and controlling
foreign trade; (ii) may prohibit, restrict and regulate exports
and imports, in all or specified cases as well as subject them
to exemptions; (iii) is authorised to formulate and announce
an export and import policy and also amend the same from
time to time, by notification in the Official Gazette; (iv) is
also authorised to appoint a 'Director General of Foreign
Trade' for the purpose of the Act, including formulation and
implementation of the export-import policy.

Accordingly, the Ministry of Commerce and Industry has

been set up as the most important organ concerned with the
promotion and regulation of foreign trade in India. In
exercise of the powers conferred by the Act, the Ministry
notifies a trade policy on a regular basis with certain
underlined objectives. The earlier trade policies were based
on the objectives of self-reliance and self-sufficiency. While,
the later policies were driven by factors like export led
growth, improving efficiency and competitiveness of the
Indian industries, etc.

With economic reforms, globalisation of the Indian economy

has been the guiding factor in formulating the trade policies.
The reform measures introduced in the subsequent policies
have focused on liberalization, openness and transparency.
They have provided an export friendly environment by
simplifying the procedures for trade facilitation. The
announcement of a new Foreign Trade Policy for a five year
period of 2004-09, replacing the hitherto nomenclature of
EXIM Policy by Foreign Trade Policy (FTP) is another step
in this direction. It takes an integrated view of the overall
development of India’s foreign trade and provides a roadmap
for the development of this sector. A vigorous export-led
growth strategy of doubling India’s share in global
merchandise trade (in the next five years), with a focus on
the sectors having prospects for export expansion and
potential for employment generation, constitute the main
plank of the policy. All such measures are expected to
enhance India's international competitiveness and aid in
further increasing the acceptability of Indian exports. The
policy sets out the core objectives, identifies key strategies,
spells out focus initiatives, outlines export incentives, and
also addresses issues concerning institutional support
including simplification of procedures relating to export
activities.
The key strategies for achieving its objectives include:-

* Unshackling of controls and creating an atmosphere of

trust and transparency;

* Simplifying procedures and bringing down transaction

costs;
 * Neutralizing incidence of all levies on inputs used in
export products;

* Facilitating development of India as a global hub for

manufacturing, trading and services;

* Identifying and nurturing special focus areas to generate

additional employment opportunities, particularly in semi-
urban and rural areas;

* Facilitating technological and infrastructural upgradation

of the Indian economy, especially through import of capital
goods and equipment;

* Avoiding inverted duty structure and ensuring that

domestic sectors are not disadvantaged in trade agreements;
 * Upgrading the infrastructure network related to the entire
foreign trade chain to international standards;

* Revitalizing the Board of Trade by redefining its role

and inducting into it experts on trade policy; and

* Activating Indian Embassies as key players in the export

strategy.

The FTP has identified certain thrust sectors having

prospects for export expansion and potential for employment
generation. These thrust sectors include: (i) Agriculture; (ii)
Handlooms & Handicrafts; (iii) Gems & Jewellery; and (iv)
Leather & Footwear. Accordingly, specific policy initiative
for these sectors have been announced.

* For the agriculture sector :-

 o A new scheme called "Vishesh Krishi Upaj Yojana
(Special Agricultural Produce Scheme)" to boost exports of
fruits, vegetables, flowers, minor forest produce and their
value added products has been introduced. Under the
scheme, exports of these products qualify for duty free credit
entitlement (5 per cent of Free On Board (f.o.b) value of
exports) for importing inputs and other goods;

o Duty free import of capital goods under Export

Promotion Capital Goods (EPCG) scheme, permitting the
installation of capital goods imported under EPCG for
agriculture anywhere in the Agri- Export Zone (AEZ);

o Utilizing funds from the 'Assistance to States for

Infrastructure Development of Exports (ASIDE) scheme' for
development of AEZs;
 o Liberalization of import of seeds, bulbs, tubers and
planting material, and liberalization of the export of plant
portions, derivatives and extracts to promote export of
medicinal plants and herbal products.

* For the handlooms and handicraft sector :-

o Enhancing to 5 per cent of Free On Board (f.o.b)

value of exports duty free import of trimmings and
embellishments for handlooms and handicrafts;

o Exemption of samples from countervailing duty

(CVD);

o Authorizing Handicraft Export Promotion Council to

import trimmings, embellishments and samples for small
manufacturers; and
 o Establishment of a new Handicraft Special Economic
Zone.

* For the gems and jewellery sector :-

o Permission for duty free import of consumables for

metals other than gold and platinum up to 2 per cent of Free
On Board (f.o.b) value of exports;

o Duty free re-import entitlement for rejected jewellery

allowed up to 2 per cent of f.o.b value of exports;

o Increase in duty free import of commercial samples

of jewellery to Rs.1 lakh; and

o Permission to import of gold of 18 carat and above

under the replenishment scheme.

* For the leather and footwear sector, the specific policy

initiatives are mainly in the form of reduction in the
incidence of customs duties on the inputs and plants and
machinery. These include:-

o Increase in the limit for duty free entitlements of

import trimmings, embellishments and footwear components
for leather industry to 3 per cent of Free On Board (f.o.b)
value of exports and that for duty free import of specified
items for leather sector to 5 per cent of f.o.b value of exports;

o Import of machinery and equipment for Effluent

Treatment Plants for leather industry exempted from customs
duty; and
 o Re-export of unsuitable imported materials (such as
raw hides and skin and wet blue leathers) has been permitted.

In order to review the progress and policy measures, each

year, "Annual Supplements" to the five year Foreign Trade
Policy (FTP) have been announced by the Ministry :-

o The Annual Supplement announced in April, 2005

incorporated additional policy initiatives and further
simplified the procedures. It provided for an active
involvement of the State Governments in creating an
enabling environment for boosting international trade, by
setting up an Inter-State Trade Council. Also, different
categories of advance licences were merged into a single
category for procedural facilitation and easy monitoring. The
supplement provided renewed thrust to agricultural exports
by extension of 'Vishesh Krish Upaj Yojna' to poultry and
dairy products and removal of cess on exports of all
agricultural and plantation commodities.

o The Annual Supplement put forward in April 2006,

announced the twin schemes of 'Focus Product' and 'Focus
Market'. To further meet the objective of employment
generation in rural and semi urban areas, export of village
and cottage industry products were included in the 'Vishesh
Krishi Upaj Yojana', which was renamed as "Vishesh Krishi
and Gram Udyog Yojana". Also, a number of measures were
introduced in order to achieve the objective of making India
a gems and jewellery hub of the world. These include:- (i)
allowing import of precious metal scrap and used jewellery
for melting, refining and re-export; (ii) permission for export
of jewellery on consignment basis; (iii) permission to export
polished precious and semi precious stones for treatment
abroad and re-import in order to enhance the quality and
afford higher value in the international market.
 o Likewise, the third Annual Supplement to the
Foreign Trade Policy was announced on 19 April,2007
(effective from 1 st April, 2007). Some of the important
measures introduced by it are:- (i) exemption from service
tax on services (related to exports) rendered abroad; (ii)
service tax on services rendered in India and utilized by
exporters would be exempted/remitted; (iii) categorization of
exporters as 'One to Five Star Export Houses' has been
changed to 'Export Houses & Trading Houses', with
rationalization and change in export performance parameters;
(iv) expansion of ceiling, scope and coverage under the
'Focus Market Scheme (FMS)' and 'Focus Product Scheme
(FPS)'.

o The final annual supplement to the Foreign Trade

Policy for 2004-2009 was announced in April 2008 in which
several innovative steps were proposed. They included the
following:

+ Import duty under the EPCG scheme is being

reduced from 5% to 3%, in order to promote modernization
of manufacturing and services exports.

+ Income tax benefit to 100% EOUs available

under Section 10B of Income Tax Act is being extended for
one more year, beyond 2009.

+ To promote export of sports and toys and also to

compensate disadvantages suffered by them, an additional
duty credit of 5% over and above the credit under 'Focus
Product Scheme' is being provided.

+ Our export of fresh fruits and vegetables and

floriculture suffers from high incidence of freight cost. To
neutralize this disadvantage, an additional credit of 2.5%
over and above the credit available under Visesh Krishi and
Gram Udyog Yojana (VKGUY) is proposed.

+ Interest relief already granted for sectors affected

adversely by the appreciation of the rupee is being extended
for one more year.

+ DEPB scheme is being continued till May 2009.

Trade Facilitation Measures (Supplement To Foreign

Trade Policy 2004-09) Announced On 26th February 2009,
o DUTY CREDIT SCRIPS under DEPB scheme to be
issued without waiting for realization of export proceeds;

o Special package of Rs.325 crore for leather and

textiles sector;

o STCL, DIAMOND INDIA, MSTC, GEM &

JEWELLERY EPC and STAR TRADING HOUSES added
as nominated agencies for import of precious metals;

o Gem and Jewellery export: import restrictions on

worked corals removed;

o Bhilwara and Surat recognized as towns of export

excellence for textiles and diamonds;

o Threshold limit for recognition as premier trading

houses reduced to Rs. 7500 crore;

o Under EPCG scheme, export obligation extended till

2009-10 for exports during 2008-09;

o DEPB/DUTY CREDIT SCRIP utilization extended

for payment of duty for import of restricted items also;
 o Procedure for claiming duty drawback refund &
refund of terminal excise duty further simplified;

o Re-credit of 4% SAD for VKGUY, FPS and FMS

allowed;

o A new office of DGFT to be opened at Srinagar;

o Value cap under DEPB revised for two products;

o Electronic message transfer facility for advance

authorization and EPCG to be established;

o Gem & Jewellery units in EOU to be allowed –

personal carriage of gold up to 10 kg;
o Advance licenses issued prior to 1.4.2002 requiring
MODVAT/CENVAT certificate dispensed with;
 o Export obligation period against advance
authorizations extended up to 36 months;

o Reimbursement of additional duty of excise levied

on fuel to be admissible for EOUS;

o Early refund of service tax claims & further

simplification of refund procedures on the anvil;

In accordance with the provisions of the Act, a

"Directorate General of Foreign Trade (DGFT)" has been set
up as an attached office of the Ministry of Commerce and
Industry. It is headed by the 'Director General of Foreign
Trade' and is responsible for formulating and executing the
Foreign Trade Policy/Exim Policy with the main objective of
promoting Indian exports. The DGFT also issues licences to
exporters and monitors their corresponding obligations
through a net work of 32 regional offices located at the
following places:- Ahmedabad; Amritsar; Bangalore; Baroda
(Vadodara); Bhopal; Kolkata; Chandigarh; Chennai;
Coimbatore; Cuttack; Ernakulam; Guwahati; Hyderabad;
Jaipur; Kanpur; Ludhiana; Madurai; Moradabad; Mumbai;
New Delhi; Panaji; Panipat; Patna; Pondicherry; Pune;
Rajkot; Shillong; Srinagar(Functioning at Jammu); Surat;
Thiruvananthapuram; Varanasi; and Vishakhapatnam.

India Exim Policy - Foreign Trade Policy.

Exim Policy or Foreign Trade Policy is a set of guidelines

and instructions established by the DGFT in matters related
to the import and export of goods in India.

The Foreign Trade

Policy of India is guided by the Export Import in known as in
short EXIM Policy of the Indian Government and is
regulated by the Foreign Trade Development and Regulation
Act, 1992.
DGFT (Directorate General of Foreign Trade) is the main
governing body in matters related to Exim Policy. The main
objective of the Foreign Trade (Development and
Regulation) Act is to provide the development and regulation
of foreign trade by facilitating imports into, and augmenting
exports from India. Foreign Trade Act has replaced the
earlier law known as the imports and Exports (Control) Act
1947.

EXIM Policy
Indian EXIM Policy contains various policy related decisions
taken by the government in the sphere of Foreign Trade, i.e.,
with respect to imports and exports from the country and
more especially
export promotion measures, policies and procedures related
thereto. Trade Policy is prepared and announced by the
Central Government (Ministry of Commerce). India's Export
Import Policy also know as Foreign Trade Policy, in general,
aims at developing export potential, improving export
performance, encouraging foreign trade and creating
favorable balance of payments position.

History of Exim Policy of India

In the year 1962, the Government of India appointed a

special

Exim Policy Committee to review the government previous

export import policies. The committee was later on approved
by the Government of India. Mr. V. P. Singh, the then
Commerce Minister and announced the Exim Policy on the
12th of April, 1985. Initially the EXIM Policy was
introduced for the period of three years with main objective
to boost the export business in India

Exim Policy Documents

The Exim Policy of India has been described in the following
documents:

* Interim New Exim Policy 2009 - 2010

* Exim Policy: 2004- 2009
* Handbook of Procedures Volume I
* Handbook of Procedures Volume II
* ITC(HS) Classification of Export- Import Items

The major information in matters related to export and

import is given in the document named "Exim Policy 2002-
2007".

An exporter uses the Handbook of Procedures Volume-I to

know the procedures, the agencies and the documentation
required to take advantage of a certain provisions of the
Indian EXIM Policy. For example, if an exporter or importer
finds out that paragraph 6.6 of the

Exim Policy is important for his export business then the

exporter must also check out the same paragraph in the
Handbook of Procedures Volume- I for further details.

The Handbook of Procedures Volume-II provides very

crucial information in matters related to the Standard Input-
Output Norms (SION). Such Input output norms are
applicable for the products such as electronics, engineering,
chemical, food products including fish and marine products,
handicraft, plastic and leather products etc. Based on SION,
exporters are provided the facility to make duty-free import
of inputs required for manufacture of export products under
the
Duty Exemption Scheme or Duty Remission Scheme.

The
Export Import Policy regarding import or export of a specific
item is given in the ITC- HS Codes or better known as
Indian Trade Clarification Code based on Harmonized
System of Coding was adopted in India for import-export
operations. Indian
Custom uses an eight digit ITC-HS Codes to suit the national
trade requirements. ITC-HS codes are divided into two
schedules. Schedule I describe the rules and
exim guidelines
related to import policies where as
Export Policy Schedule II describe the rules and regulation
related to export policies. Schedule I of the ITC-HS code is
divided into 21 sections and each section is further divided
into chapters. The total number of chapters in the schedule I
is 98. The chapters are further divided into sub-heading
under which different HS codes are mentioned.

ITC(Hs) Schedule II of the code contain 97 chapters giving

all the details about the
Export Import Guidelines related to the export policies.

Objectives Of The Exim Policy : -

Government control import of non-essential items through

the

EXIM Policy. At the same time, all-out efforts are made to

promote exports. Thus, there are two aspects of Exim Policy;
the import policy which is concerned with regulation and
management of imports and the export policy which is
concerned with exports not only promotion but also
regulation. The main objective of the Government's EXIM
Policy is to promote exports to the maximum extent. Exports
should be promoted in such a manner that the economy of
the country is not affected by unregulated exportable items
specially needed within the country. Export control is,
therefore, exercised in respect of a limited number of items
whose supply position demands that their exports should be
regulated in the larger interests of the country. In other
words, the main objective of the Exim Policy is:

* To accelerate the economy from low level of economic

activities to high level of economic activities by making it a
globally oriented vibrant economy and to derive maximum
benefits from expanding global market opportunities.
* To stimulate sustained economic growth by providing
access to essential raw materials, intermediates, components,'
consumables and capital goods required for augmenting
production.
* To enhance the techno local strength and efficiency of
Indian agriculture, industry and services, thereby, improving
their competitiveness.
* To generate new employment.
* Opportunities and encourage the attainment of
internationally accepted standards of quality.
* To provide quality consumer products at reasonable
prices.

Governing Body of Exim Policy

The Government of India notifies the Exim Policy for a

period of five years (1997-2002) under Section 5 of the
Foreign Trade (Development and Regulation Act), 1992. The
current
Export Import Policy covers the period 2002-2007. The
Exim Policy is updated every year on the 31st of March and
the modifications, improvements and new schemes became
effective from 1st April of every year.

All types of changes or modifications related to the EXIM

Policy is normally announced by the Union Minister of
Commerce and Industry who co-ordinates with the Ministry
of Finance, the
Directorate General of Foreign Trade
and network of
Dgft Regional
Offices.

Exim Policy 1992 -1997

In order to
liberalize imports and boost exports, the Government of
India for the first time introduced the Indian Exim Policy on
April I, 1992. In order to bring stability and continuity, the
Export Import Policy was made for the duration of 5 years.
However, the Central Government reserves the right in
public interest to make any amendments to the trade Policy
in exercise of the powers conferred by Section-5 of the Act.
Such amendment shall be made by means of a Notification
published in the Gazette of India.
Export Import Policy is believed to be an important step
towards the economic reforms of India.

Exim Policy 1997 -2002

With time the Exim Policy 1992-1997 became old, and a

New Export Import Policy was need for the smooth
functioning of the Indian export import trade. Hence, the
Government of India introduced a new Exim Policy for the
year 1997-2002. This policy has further simplified the
procedures and educed the interface between exporters and
the Director General of Foreign Trade (DGFT) by reducing
the number of documents required for export by half. Import
has been further liberalized and better efforts have been made
to promote Indian exports in international trade.

Objectives of the Exim Policy 1997 -2002

The principal objectives of the Export Import Policy 1997

-2002 are as under:

* To accelerate the economy from low level of economic

activities to high level of economic activities by making it a
globally oriented vibrant economy and to derive maximum
benefits from expanding global market opportunities.
* To motivate sustained economic growth by providing
access to essential raw materials, intermediates, components,'
consumables and capital goods required for augmenting
production.
To improve the technological strength and efficiency of
Indian agriculture, industry and services, thereby, improving
their competitiveness.
* To create new employment. Opportunities and
encourage the attainment of internationally accepted
standards of quality.
* To give quality consumer products at practical prices.

Highlights of the Exim Policy 1997-2002

1. Period of the Exim Policy

• This policy is valid for five years instead of three years as

in the case of earlier policies. It is effective from 1st April
1997 to 31st March 2002.
 2. Liberalization

• A very important feature of the policy is liberalization.

• It has substantially eliminated licensing, quantitative
restrictions and other regulatory and discretionary controls.
All goods, except those coming under negative list, may be
freely imported or exported.

3. Imports Liberalization

• Of 542 items from the restricted list 150 items have been
transferred to Special Import Licence (SIL) list and
remaining 392 items have been transferred to Open General
Licence (OGL) List.

4. Export Promotion Capital Goods (EPCG) Scheme

• The duty on imported capital goods under

EPCG Scheme has been reduced from 15% to 10%.
• Under the zero duty EPCG Scheme, the threshold limit has
been reduced from Rs. 20 crore to Rs. 5 crore for agricultural
and allied Sectors

5. Advance Licence Scheme

• Under Advance License Scheme, the period for export

obligation has been extended from 12 months to 18 months.
• A further extension for six months can be given on payment
of 1 % of the value of unfulfilled exports.

6. Duty Entitlement Pass Book (DEPB) Scheme

• Under the DEPB

Scheme an exporter may apply for credit, as a specified
percentage of FOB value of exports, made in freely
convertible currency.
• Such credit can be can be
utilized for import of raw materials, intermediates,
components, parts, packaging materials, etc. for export
purpose.

Impact of Exim Policy 1997 –2002

(a) Globalization of Indian Economy:

The Exim Policy 1997-02 proposed with an aim to prepare a
framework for globalizations of Indian economy. This is
evident from the very first objective of the policy, which
states. "To accelerate the economy from low level of
economic activities to- high level of economic activities by
making it a globally oriented vibrant economy and to derive
maximum benefits from expanding global market
opportunities."
(b) Impact on the Indian Industry:
In the EXIM policy 1997-02, a series of reform measures
have been introduced in order to give boost to India's
industrial growth and generate employment opportunities in
non-agricultural sector. These include the reduction of duty
from 15% to 10% under EPCG scheme that enables Indian
firms to import capital goods and is an important step in
improving the quality and productivity of the Indian industry.

(c) Impact on Agriculture:

Many encouraging steps have been taken in the Exim Policy
1997-2002 in order to give a boost to Indian agricultural
sector. These steps includes provision of additional SIL of 1
% for export of agro products, allowing EOU’s and other
units in EPZs in agriculture sectors to 50% of their output in
the domestic tariff area (DTA) on payment of duty.

(d) Impact on Foreign Investment.

In order to encourage foreign investment in India, the Exim
Policy 1997-02 has permitted 100% foreign equity
participation in the case of 100% EOUs, and units set up in
EPZs.

(e) Impact on Quality up gradation:

The SIL entitlement of exporters holding ISO 9000
certification has been increased from 2% to 5% of the FOB
value of exports, which has encouraged Indian industries to
undertake research and development programmers and
upgrade the quality of their products.

(f) Impact on Self-Reliance:-

The Exim Policy 1997-2002 successfully fulfills one of the
India’s long terms objective of Self-reliance. The Exim
Policy has achieved this by encouraging domestic sourcing
of raw materials, in order to build up a strong domestic
production base. New incentives added in the Exim Policy
have also added benefits to the exporters.

Exim Policy 2002 – 2007

The Exim Policy 2002 - 2007 deals with both the export and
import of merchandise and services. It is worth mentioning
here that the Exim Policy: 1997 - 2002 had accorded a status
of exporter to the business firm exporting services with effect
from1.4.1999. Such business firms are known as Service
Providers.

Objectives of the Exim Policy: 2002 - 2007

The main objectives of the Export Import Policy 2002-2007
are as follows:

1. To encourage economic growth of India by providing

supply of essential raw materials, intermediates, components,
consumables and capital goods required for augmenting
production and providing services.
 2. To improve the technological strength and efficiency of
Indian agriculture, industry and services, thereby improving
their competitive strength while generating new employment
opportunities and encourage the attainment of internationally
accepted standards of quality; and
3. To provide consumers with good quality products and
services at internationally competitive prices while at the
same time creating a level playing field for the domestic
producers.

Main Elements of Exim Policy 2004-2009

The new Exim Policy 2004-2009 has the following main

elements:

* Preamble
* Legal Framework
* Special Focus Initiatives
 * Board Of Trade
* General Provisions Regarding Imports And Exports
* Promotional Measures
* Duty Exemption / Remission Schemes
* Export Promotion Capital Goods Scheme
* Export Oriented Units (EOUs),Electronics Hardware
Technology Parks (EHTPS), Software Technology Parks
(STPs) and Bio-Technology Parks (BTPs)
* Special Economic Zones
* Free Trade & Warehousing Zones
* Deemed Exports

Permeable of Exim Policy 2004-2009: It is a speech given by

the Ministry of Commerce and Industries. The speech for the
Exim Policy 2004-2009 was given by Kamal Nath, on 31ST
AUGUST, 2004.

Legal Framework of Exim Policy 2004-2009

1.1 Preamble
The Preamble spells out the broad framework and is an
integral part of the Foreign Trade Policy.

1.2 Duration
In exercise of the powers conferred under Section 5 of The
Foreign Trade (Development and Regulation Act), 1992 (No.
22 of 1992), the Central Government hereby notifies the
Exim Policy for the period 2004-2009 incorporating the
Export Import Policy for the period 2002-2007, as modified.
This Policy shall come into force with effect from 1st
September, 2004 and shall remain in force up to 31st March,
2009, unless as otherwise specified.

1.3 Amendments
The Central Government reserves the right in public interest
to make any amendments to this Policy in exercise of the
powers conferred by Section-5 of the Act. Such amendment
shall be made by means of a Notification published in the
Gazette of India.

1.4 Transitional Arrangements

Notifications made or Public Notices issued or anything done
under the previous Export / Import policies and in force
immediately before the commencement of this Policy shall,
in so far as they are not inconsistent with the provisions of
this Policy, continue to be in force
and shall be deemed to have been made, issued or done under
this Policy.

Licenses, certificates and permissions issued before the

commencement of this Policy shall continue to be valid for
the purpose and duration for which such licence; certificate
or permission was issued unless otherwise stipulated.
1.5 Free Export Import
In case an export or import that is permitted freely under
Export Import Policy is subsequently subjected to any
restriction or regulation, such export or import will ordinarily
be permitted notwithstanding such restriction or regulation,
unless otherwise stipulated, provided that the shipment of the
export or import is made within the original validity of an
irrevocable letter of credit established before the date of
imposition of such restriction.

Special Focus Initiative of Exim Policy 2004-2009

With a view to doubling our percentage share of global trade
within 5 years and expanding employment opportunities,
especially in semi urban and rural areas, certain special focus
initiatives have been identified for agriculture, handlooms,
handicraft, gems & jewellery, leather and Marine sectors.

Government of India shall make concerted efforts to promote

exports in these sectors by specific sectoral strategies that
shall be notified from time to time.

Board of Trade of Exim Policy 2004-2009

BOT has a clear and dynamic role in advising government on

relevant issues connected with foreign trade.

* To advise Government on Policy measures for

preparation and implementation of both short and long term
plans for increasing exports in the light of emerging national
and international economic scenarios;
* To review export performance of various sectors,
identify constraints and suggest industry specific measures to
optimize export earnings;
* To examine existing institutional framework for imports
and exports and suggest practical measures for further
streamlining to achieve desired objectives;
 * To review policy instruments and procedures for imports
and exports and suggest steps to rationalize and channelize
such schemes for optimum use;
* To examine issues which are considered relevant for
promotion of India’s foreign trade, and to strengthen
international competitiveness of Indian goods and services;
and
* To commission studies for furtherance of above
objectives.

General Provisions Regarding Exports and Imports of Exim

Policy 2004-2009
The Export Import Policy relating to the general provisions
regarding exports and Imports is given in Chapter-2 of the
Exim Policy.

Countries of Imports/Exports - Unless otherwise specifically

provided, import/ export will be valid from/to any country.
However, import/exports of arms and related material
from/to Iraq shall be prohibited.

The above provisions shall, however, be subject to all

conditionality, or requirement of licence, or permission, as
may be required under Schedule II of ITC (HS).

Promotional Measures of Exim Policy 2004-2009

The Government of India has set up several institutions

whose main functions are to help an exporter in his work. It
would be advisable for an exporter to acquaint him with
these institutions and the nature of help that they can provide
so that he can initially contact them and have a clear picture
of what help he can expect of the organized sources in his
export effort. Some of these institution are as follows.

Export Promotion Councils

Commodity Boards
Marine Products Export Development Authority

Agricultural & Processed Food Products Export

Development Authority
Indian Institute of Foreign Trade
India Trade Promotion Organization (ITPO)
National Centre for Trade Information (NCTI)
Export Credit Guarantee Corporation (ECGC)
Export-Import Bank
Export Inspection Council
Indian Council of Arbitration
Federation of Indian Export
Organizations
Department of Commercial Intelligence and Statistics
Directorate General of Shipping
Freight Investigation Bureau
Duty Exemption / Remission Schemes of Exim Policy 2004-
2009
The Duty Exemption Scheme enables import of inputs
required for export production. It includes the following
exemptions-

Duty Drawback: - The Duty Drawback Scheme is

administered by the Directorate of Drawback, Ministry of
Finance. Under Duty Drawback scheme, an exporter is
entitled to claim
Indian Customs Duty paid on the imported goods and Central
Excise Duty paid on indigenous raw materials or
components.

Excise Duty Refund: - Excise Duty is a tax imposed by the

Central Government on goods manufactured in India. Excise
duty is collected at source, i.e., before removal of goods from
the factory premises. Export goods are totally exempted from
central excise duty.

Octroi Exemption: - Octroi is a duty paid on manufactured

goods, when they enter the municipal limits of a city or a
town. However, export goods are exempted from Octroi.

The Duty Remission Scheme enables post export

replenishment/ remission of duty on inputs used in the export
product.

DEPB: Duty Entitlement Pass Book in short DEPB

Rate is basically an export incentive scheme. The objective
of DEPB Scheme is to neutralize the incidence of basic
custom duty on the import content of the exported products.

DFRC
Under the Duty Free Replenishment Certificate (DFRC)
schemes, import incentives are given to the exporter for the
import of inputs used in the manufacture of goods without
payment of basic customs duty. Duty Free Replenishment
Certificate (DFRC) shall be available for exports only up to
30.04.2006 and from 01.05.2006 this scheme is being
replaced by the

Duty Free Import Authorisation (DFIA).

DFIA: Effective from 1st May, 2006, Duty Free Import

Authorisation or DFIA in short is issued to allow duty free
import of inputs which are used in the manufacture of the
export product (making normal allowance for wastage), and
fuel, energy, catalyst etc. which are consumed or utilised in
the course of their use to obtain the export product. Duty
Free Import Authorisation is issued on the basis of inputs and
export items given under Standard Input and Output
Norms(SION).
Export Oriented Units (EOUs), Electronics Hardware
Technology Parks (EHTPs), Software Technology
Parks(STPs) And Bio-Technology Parks (BTPs) of Exim
Policy 2004-2009
The Export Import Policies relating to Export Oriented Units
(EOUs) Electronics Hardware Technology Parks (EHTPs),
Software Technology Parks (STPs) and Bio-technology
parks (BTPs) Scheme is given in Chapter 6 of the Foreign
Trade Policy. Software Technology Park(STP)/Electronics
Hardware Technology Park (EHTP) complexes can be set up
by the Central Government, State Government, Public or
Private Sector Undertakings.

Export Promotion Capital Goods Scheme (EPCG) of Exim

Policy 2004-2009
Introduced in the EXIM policy of 1992-97,
Export Promotion Capital Goods Scheme (EPCG) enable
exporters to import machinery and other capital goods for
export production at concessional or no customs duties at all.
This facility is subject to export obligation, i.e., the exporter
is required to guarantee exports of certain minimum value,
which is in multiple of total value of capital goods imported.

Capital goods imported under EPCG Scheme are subject to

actual user condition and the same cannot be transferred
/sold till the fulfillment of export obligation specified in the
licence. In order to ensure that the capital goods imported
under EPCG Scheme, the licence holder is required to
produce certificate from the jurisdictional
Central Excise Authority (CEA) or Chartered Engineer (CE)
confirming installation of such capital goods in the declared
premises.

Special Economic Zone (SEZ)

under the Exim Policy 2004-2009
A Special Economic Zone in short SEZ is a geographically
distributed area or zones where the economic laws are more
liberal as compared to other parts of the country. SEZs are
proposed to be specially delineated duty free enclaves for the
purpose of trade, operations, duty and tariffs. SEZs are self-
contained and integrated having their own infrastructure and
support services.

The area under 'SEZ' covers a broad range of zone types,

including Export Processing Zones (EPZ), Free Zones (FZ),
Industrial Estates (IE), Free Trade Zones (FTZ), Free Ports,
Urban Enterprise Zones and others.

In Indian, at present there are eight functional Special

Economic Zones located at Santa Cruz (Maharashtra),
Cochin (Kerala), Kandla and Surat (Gujarat), Chennai (Tamil
Nadu), Visakhapatnam (Andhra Pradesh), Falta (West
Bengal) and Noida (Uttar Pradesh) in India. Further a Special
Economic Zone at Indore ( Madhya Pradesh ) is also ready
for operation.

Free Trade & Warehousing Zones of Exim Policy 2004-2009

Free Trade & Warehousing Zones (FTWZ) shall be a special
category of Special Economic Zones with a focus on trading
and warehousing. The concept of FTWZ is new and has been
recently introduced in the five-year foreign trade policy
2004-09. Its main objective is to provide infrastructure for
growth of the economy and foreign trade. Free Trade &
Warehousing Zones (FTWZ) plays an important role in
achieving global standard warehousing facilities as free trade
zones. Free Trade & Warehousing Zones is a widely
accepted model with a history of providing Substantial
encouragement to foreign trade and warehousing activity.

Deemed Exports under the Exim Policy 2004-2009

Deemed Export is a special type of transaction in the Indian
Exim policy in which the payment is received before the
goods are delivered. The payment can be done in Indian
Rupees or in Foreign Exchange. As the deemed export is also
a source of foreign exchange, so the Government of India has
given the benefit duty free import of inputs.

What is the WTO?

The World Trade Organization (WTO) is the only global

international organization dealing with the rules of trade
between nations. At its heart are the WTO agreements,
negotiated and signed by the bulk of the world’s trading
nations and ratified in their parliaments. The goal is to help
producers of goods and services, exporters, and importers
conduct their business
The World Trade Organization (WTO) is an international
organization designed by its founders to supervise and
liberalize international trade. The organization officially
commenced on January 1, 1995 under the Marrakech
Agreement, replacing the General Agreement on Tariffs and
Trade (GATT), which commenced in 1947.

The World Trade Organization deals with regulation of trade

between participating countries; it provides a framework for
negotiating and formalising trade agreements, and a dispute
resolution process aimed at enforcing participants' adherence
to WTO agreements which are signed by representatives of
member governments and ratified by their parliaments.[4][5]
Most of the issues that the WTO focuses on derive from
previous trade negotiations, especially from the Uruguay
Round (1986-1994).

The organization is currently endeavouring to persist with a

trade negotiation called the Doha Development Agenda (or
Doha Round), which was launched in 2001 to enhance
equitable participation of poorer countries which represent a
majority of the world's population. However, the negotiation
has been dogged by "disagreement between exporters of
agricultural bulk commodities and countries with large
numbers of subsistence farmers on the precise terms of a
'special safeguard measure' to protect farmers from surges in
imports. At this time, the future of the Doha Round is
uncertain."[6]

The WTO has 153 members,[7] representing more than 97%

of total world trade[8] and 30 observers, most seeking
membership. The WTO is governed by a ministerial
conference, meeting every two years; a general council,
which implements the conference's policy decisions and is
responsible for day-to-day administration; and a director-
general, who is appointed by the ministerial conference. The
WTO's headquarters is at the Centre William Rappard,
Geneva, Switzerland.

Functions

Among the various functions of the WTO, these are regarded

by analysts as the most important:

* It oversees the implementation, administration and

operation of the covered agreements.[28][29]
* It provides a forum for negotiations and for settling
disputes.[30][31]

Additionally, it is the WTO's duty to review and propagate

the national trade policies, and to ensure the coherence and
transparency of trade policies through surveillance in global
economic policy-making.[29][31] Another priority of the
WTO is the assistance of developing, least-developed and
low-income countries in transition to adjust to WTO rules
and disciplines through technical cooperation and training.
[32]

The WTO is also a center of economic research and analysis:

regular assessments of the global trade picture in its annual
publications and research reports on specific topics are
produced by the organization.[33] Finally, the WTO
cooperates closely with the two other components of the
Bretton Woods system, the IMF and the World Bank.[3
About the WTO

The World Trade Organization (WTO), established on 1

January 1995, is the legal and institutional foundation of the
multilateral trading system. It provides the principal
contractual obligations determining how governments frame
and implement domestic trade legislation and regulations.
And it is the platform on which trade relations among
countries evolve through collective debate, negotiation and
adjudication.

The WTO is the embodiment of the results of the Uruguay

Round trade negotiations and the successor to the General
Agreement on Tariffs and Trade (GATT).

Objectives

The Preamble of the Agreement Establishing the WTO states

that members should conduct their trade and economic
relations with a view to "raising standards of living, ensuring
full employment and a large and steadily growing volume of
real income and effective demand, and expanding the
production of and trade in goods and services, while allowing
for the optimal use of the world's resources in accordance
with the objective of sustainable development, seeking both
to protect and preserve the environment and to enhance the
means for doing so in a manner consistent with their
respective needs and concerns at different levels of
development."

Furthermore, members recognize the "need for positive

efforts designed to ensure that developing countries, and
especially the least-developed among them, secure a share in
international trade commensurate with the needs of their
economic development."

To contribute to the achievement of these objectives, WTO

Members have agreed to enter into "reciprocal and mutually
advantageous arrangements directed to the substantial
reduction of tariffs and other barriers to trade and to the
elimination of discriminatory treatment in international trade
relations."

Key Principles
The fundamental principles of the multilateral trading system
are:

- Trade without discrimination. Under the "most-favoured

nation" (MFN) clause, members are bound to grant to the
products of other members no less favourable treatment than
that accorded to the products of any other country. The
provision on "national treatment" requires that once goods
have entered a market, they must be treated no less
favourably than the equivalent domestically-produced good.

- Predictable and growing access to markets. While quotas

are generally outlawed, tariffs or customs duties are legal in
the WTO. Tariff reductions made by over 120 countries in
the Uruguay Round are contained in some 22,500 pages of
national tariff schedules which are considered an integral part
of the WTO. Tariff reductions, for the most part phased in
over five years, will result in a 40 per cent cut in industrial
countries' tariffs in industrial products from an average of 6.3
per cent to 3.8 per cent. The Round also increased the
percentage of bound product lines to nearly 100 per cent for
developed nations and countries in transition and to 73 per
cent for developing countries. Members have also undertaken
an initial set of commitments covering national regulations
affecting various services activities. These commitments are,
like those for tariffs, contained in binding national schedules.

- Promoting fair competition. The WTO extends and clarifies

previous GATT rules that laid down the basis on which
governments could impose compensating duties on two
forms of "unfair" competition: dumping and subsidies. The
WTO Agreement on agriculture is designed to provide
increased fairness in farm trade. That on intellectual property
will improve conditions of competition where ideas and
inventions are involved, and another will do the same thing
for trade in services.
- Encouraging development and economic reform. GATT
provisions intended to favour developing countries are
maintained in the WTO, in particular those encouraging
industrial countries to assist trade of developing nations.
Developing countries are given transition periods to adjust to
the more difficult WTO provisions. Least-developed
countries are given even more flexibility and benefit from
accelerated implementation of market access concessions for
their goods.

Main functions

The essential functions of the WTO are:

- administering and implementing the multilateral and

plurilateral trade agreements which together make up the
WTO;
- acting as a forum for multilateral trade negotiations;

- seeking to resolve trade disputes;

- overseeing national trade policies; and

- cooperating with other international institutions involved in

global economic policy-making.

Structure

The highest WTO authority is the Ministerial Conference

which meets every two years. The day-to-day work of the
WTO, however, falls to a number of subsidiary bodies,
principally the General Council, which also convenes as the
Dispute Settlement Body and as the Trade Policy Review
Body. The General Council delegates responsibility to three
other major bodies - namely the Councils for Trade in
Goods, Trade in Services and Trade-Related Aspects of
Intellectual Property Rights.

Three other bodies are established by the Ministerial

Conference and report to the General Council: the
Committee on Trade and Development, the Committee on
Balance of Payments and the Committee on Budget, Finance
and Administration. The General Council formally
established, in early 1995, a Committee on Trade and
Environment, which will present a report on its work to the
first meeting of the WTO Ministerial Conference, scheduled
for December 1996 in Singapore.

Each of the plurilateral agreements of the WTO - those on

civil aircraft, government procurement, dairy products and
bovine meat - have their own management bodies which
report to the General Council.
Secretariat and budget

The WTO Secretariat is located in Geneva. It has around 450

staff and is headed by its Director-General, Renato Ruggiero,
and four Deputy Directors-General. Its responsibilities
include the servicing of WTO delegate bodies with respect to
negotiations and the implementation of agreements. It has a
particular responsibility to provide technical support to
developing countries, and especially the least-developed
countries. WTO economists and statisticians provide trade
performance and trade policy analyses while its legal staff
assist in the resolution of trade disputes involving the
interpretation of WTO rules and precedents. Other
Secretariat work is concerned with accession negotiations for
new members and providing advice to governments
considering membership.
The WTO budget is around US$83 million with individual
contributions calculated on the basis of shares in the total
trade conducted by members. Part of the budget also goes to
the International Trade Centre.

Data Preparation, Interpretation and Analysis

Analyzing survey data is an important and exciting step in

the survey process. It is the time that you may reveal
important facts about your customers, uncover trends that
you might not otherwise have known existed, or provide
irrefutable facts to support your plans. By doing in-depth
data comparisons, you can begin to identify relationships
between various data that will help you understand more
about your respondents, and guide you towards better
decisions.
This article gives you a brief overview of how to analyze
survey results. It does not discusses specific usage of
eSurveysPro for conducting analysis as it is intended to
provide a foundation upon which you can confidently
conduct your own survey analysis no matter what tool you
use.
Three Common Mistakes

Before you dive into analyzing your survey results, take a

look back at the big picture. What objectives were you trying
to accomplish when you created your survey? Did your
survey instrument meet those objectives? Is the data you
collected the right data? Do you have sufficient data to
properly reach a conclusion?

Although data analysis is the wrong time to try and rewrite

your survey instrument, it is important to remember the
scope of your project and stick to it. Many first time
surveyors attempt to read "between the lines" while
analyzing data. They attempt to answer questions that were
not asked by making inferences and assumptions from those
that were asked. Doing so amounts to nothing more than
guesswork. To avoid this temptation, remember this simple
rule:
Rule 1: If you did not ask you do not know.

Another common mistake that many first time surveyors

make is to attempt to change data to compensate for poor
question design. For example, if a question asked a
respondent to indicate his total household income using a
scale of values, a mean and median cannot be calculated.
Many people try to get around this by assigning each
response a value representing the range. Even if the
adjustment is made consistently across all responses, the
resulting calculations will be wrong. Similarly, trying to
analyze a multiple-choice question as if it was a single-select
question will often provide erroneous information. In order
to avoid this pitfall, remember this simple rule:
Rule 2: Do not alter data to compensate for bad survey
design.

A second mistake inexperienced surveyors make is to project

the findings to an audience that was not either part of the
survey population or not adequately represented. For
example, if an HR manager conducts a benefits survey and
invites all employees to participate, most people would
assume that the results represent all employees since
everyone had an opportunity to participate. Provided that
enough employees participate, the data might be statistically
valid, but is it really representative of all employees? The
answer is, it depends. If the survey collected data about
employee demographics that could be compared to what is
known about the company, then the results do reflect the
company as a whole. However, if 80% of the respondents are
married and 50% of the total employee base is married, the
results of the survey are skewed toward married people. If
married people have different benefits needs than single
people, using the survey results to make conclusions about
the entire employee pool would be less accurate than those
conclusions about the married employees or single
employees independently. To avoid this temptation,
remember this simple rule:
Rule 3: Do not project your data to people that did not
respond.

The earlier you recognize flaws in your survey design and

data collection, the more time you will save during analysis.
If you questions do not provided the data you need to meet
your survey objectives, you'll have to start over. If your
questions are vague or ambiguous, you'll have to throw them
out. If you do not have an adequate number of responses,
you'll have to get more.
Survey Analysis

Analyzing any survey, web or traditional, consists of a

number of interrelated processes that are intended to
summarize, arrange, and transform data into information. If
your survey objective was simply to collect data for your
database or data warehouse, you do not have to do any
analysis of the data. On the other hand, if your objective was
to understand the characteristics of typical customers, then
you must transform you raw results in to information that
will enable you to paint a clear picture of your customers.

Assuming you need to analyze the data collected from your

survey, the process begins with a quick review of the results,
followed by editing, analysis, and reporting. To ensure you
have accurate data before investing significant time in
analysis, it is important that you do not begin analyzing
results until you have completed the review and editing
process.
Quick Review

Read all your results. Although, this seems like an obvious

thing to do, many surveyors think that they can skip this step
and dive right in to data analysis. A quick review can tell you
lots about your project, including any flaws in questionnaire
design or response population, before you spend hours of
time in analyzing the data.

During the quick review, you should look at every question

and see if the results "make sense". This "gut feel" check of
the data will often uncover any issues with your survey
project. Most surveyors already have an idea of how they
expect their data to look. A quick review of the data can help
you quickly understand that tell you if the people that
respond are the right people. For example, if you were
conducting a survey of all the employees in a company and
you knew that 10% were in the marketing department, 20%
in sales, 45% in manufacturing, 5% in management, and 5%
finance, and 15% research and development, you could
reasonable expect your responses to be similarly distributed.
If your quick review disclosed 80% of your respondents were
from the sales department, you know that your survey did not
adequately capture a representative sample of all departments
within the company.

The quick review can also highlight any problems with the
survey instrument. Are most respondents answering all
questions? If not, your questionnaire could be flawed in such
a way that a person cannot complete the survey. A low
response rate could mean your survey invitation was not
compelling enough to encourage participation, or your timing
was off and a follow-up reminder is needed.
Lastly, the quick review of the survey can show you what
areas to focus on for detailed analysis. As stated earlier, most
surveyors already know what they expect to get, so your
quick review can show you the unexpected.
Editing and Cleaning

Editing and cleaning data is an important step in the survey

process. Special care must be taken when editing survey data
so that you do not alter or throw out responses in such a way
as to bias your results. Although you can begin editing and
cleaning your data as soon as results are received, caution
should be used since any edits can be lost if the database is
rebuilt. To be safe, wait until all data is received before you
begin the editing and cleaning process.

To start, find and delete incomplete and duplicate responses.

A response should be discarded if the respondent did not
complete enough of the survey to be meaningful. For
example, if a your survey was intended to determine future
buying intentions across various demographic groups and the
respondent did not answer any of the demographic questions,
you should delete the response. On the other hand, if the
respondent answered all the demographic questions but
omitted their name or email address, then you should keep
the response.

Duplicate responses are a unique issue for electronic surveys.

Many tools, such as eSurveysPro, provide built in features to
help minimize the risk of duplicate responses. Others, like
the popular "infotainment" polls featured on many websites
do nothing to eliminate duplicates. Without removing
duplicates, your data will be skewed in favor of the duplicate
response. Both the count and percentage of the whole will be
affected by duplicate responses, and computed means and
medians will also be thrown off. To find duplicate responses,
carefully examine the answers to any open-ended questions.
When two open-ended questions have the exact same
answer, a duplicate response is likely to exist. Make sure the
response is indeed a duplicate by comparing the answers to
all the other questions, and then delete one of the responses if
a match is found.

Data cleaning of web surveys usually involves categorizing

answers to open-ended questions and multiple-choice
questions that include an "other, please specify" response.
Because of their nature, open-ended text response questions
can provide significant value but they are nearly impossible
to process without some form of summarization or
tabulation. One of the easiest ways to summarize these
questions is to build a list of themes and select the themes
that apply as you read each response. Tools such as
eSurveysPro allow you to add questions after a survey is run
to do just this sort of thing.
A common problem in any survey that needs attention during
the editing and cleaning process is when a respondent
answers an "other, please specify" question by selecting
"other" and then writing in an answer that was one of the
listed response options. Without cleaning these answers, the
"other" response will be overstated and the correct response
will be understated. For example, a demographics question
that asks for the respondent's role within the organization
may have a response like "faculty, teacher, or student" and a
respondent selects "other" and types "professor," you would
want to clean the response by switching the other choice to
the one for "faculty, teacher, or student".

Once the data preparation is complete, it is time to start

analyzing the data and turning it into actionable information.
Detailed Analysis

Analysis is the most important aspect of your survey research

project. At this point, you have collected a set of data that
must now be turned into actionable information. The process
of analysis can lead to a variety of alternative courses of
action. Mistakes during analysis can lead to costly decisions
down the road, so extreme caution and careful review must
be followed throughout the process. Carelessness during
analysis can lead to disaster. What you do during analysis
will ultimately determine if your survey project is a
successful or not.

Depending on what type of information you are trying to

know about your audience, you will have to decide what
analysis makes sense. It can be as simple as reviewing the
graphs that eSurveysPro automatically creates, or conducting
in-depth comparisons between questions sets to identify
trends or relationships. For most surveyors, a basic analysis
using charts, cross tabulations, and filters is sufficient. On the
other hand, more sophisticated users may wish to do a more
complex statistical analysis using high powered analytical
tools such as SPSS, Excel, or any number of number
crunching applications. For our purposes in this article, we
will focus on basic analysis techniques.
Graphical Analysis

Graphical analysis simply means displaying the data in a

variety of visual formats that make it easy to see patterns and
identify differences among the results set. There are many
different graphing options available to display data, the most
common are Bar, Pie, and Line charts.

Bar charts use solid bars on an X and Y-axis that extend to

meet a specific data value indicated on the chart and can be
shown either vertically or horizontally. These charts are
flexible and are most commonly used to display data from
multiple-select, rank order, single-select matrix and
numerical questions. Each response option is shown as an
independent bar on the chart, and the length of the bar
represents the frequency the response was chosen relative to
all choices.

Pie charts, or circle graphs, have colorful "slices"

representing segments of your data. These charts measure
values as compared to a "whole", and the total percentages of
the segments always add up to 100%. Pie charts are most
useful with single-select questions because the each response
is represented visually as a portion of the entire pie. It is easy
to interpret which answer received the most responses in a
pie chart by selecting the largest potion of the pie. When
comparing two sets of data using a pie chart, it is important
to make sure the colors used for each response option remain
consistent in each chart. If represent the same response
options in each chart, this way, a side-by-side visual
comparison can quickly be made. Pie charts are not
appropriate for multiple-select questions because each
respondent can answer choose more than one option, and the
sum of the option percentages will exceed 100%.

There are other graphing options such as line charts, area

charts and scatter graphs, which are useful when displaying
the same data over a period of time. However these formats
are not as easy to interpret for casual users, so they should be
used sparingly.
Frequency Tables

Frequency tables are another form of basic analysis. These

tables show the possible responses, the total number of
respondents for each part, and the percentages of respondents
who selected each answer. Frequency tables are useful when
a large number of response options are available, or the
differences between the percentages of each option are small.
In most cases, pie or bar charts are easier to work with than
frequency tables.
Response Count Percent
Market Analysis 76 13.7%
Quantitative Analysis 150 27.0%
Strategic Planning 56 10.1%
Product Planning 33 5.9%
Promotional Communication 243 43.8%
Creating sales tools 152 27.4%
Providing channel support 157 28.3%
Cross Tabulation

Cross tabulations, or cross tabs, are a good way to compare

two subgroups of information. Cross tabs allow you to
compare data from two questions to determine if there is a
relationship between them. Like frequency tables, cross tabs
appear as a table of data showing answers to one question as
a series of rows and answers to another question as a series
of columns.
Base Question Female Male
Product Manager 57.2% 53.4%
Director 12.6% 14.2%
Product Marketing Manager 24.7% 23.1%
Program Manager 2.8% 1.5%
Technical Product Manager 2.8% 7.7%
Total Counts 215 337

Cross tabs are used most frequently to look at answers to a

question among various demographic groups. The
intersections of the various columns and rows, commonly
called cells, are the percentages of people who answered
each of the responses. In the example above, females and
males had relatively similar distribution among various job
titles, with the exception of the tile of "Technical Product
Manager", where 2.5 times as many males had the title as
compared to females. For analysis purposes, cross tabs are a
great way to do comparisons.
Filtering
Filtering is the most under-utilized tool used in analysis.
Filters allow you select specific subsets of data to view.
Unlike a cross tab, that compares two questions, a filter will
allow you to examine all questions for a particular subset of
the responses. By viewing only the data from the people who
responded negatively, look at how they answered other
questions. Find patterns or trends that help define why a
person answered the way they did. You can even filter on
multiple questions and criteria to do a more detailed search if
necessary. For example, if you wanted to know the buying
intentions of men, over the age of 40, with income of about
$50,000, you would set a filter that would remove all those
respondents that do not meet your criteria from the results
set, thus enabling you to concentrate on the target population.

By applying filters to the date survey responses were

received, you can see how the answers change from one time
frame to the next. For instance, by continually running a
customer satisfaction survey, you can assess changes in
customer attitudes over time by filtering on the date the
survey was received. You can also use a filter on date
received to assess the impact of sales incentive programs or
new product offerings by comparing survey responses before
and after the change.

Filters do not permanently remove the responses of those

people that do not match the specified criteria; they simply
eliminate them from the current view of the data, making it
much easier to perform analysis. By looking at the same
question with different filters applied, differences between
the various respondents represented by the filter can be
quickly seen. Because filters remain in effect until cleared,
don't forget to clear them before attempting to analyze your
survey responses as a whole, otherwise your observations
will be inaccurate, and your recommendations flawed.
Simple Regression Analysis

Determining what factors have lead to a particular outcome

is called regression analysis. The regression means you're
working backwards from the result to find out why a person
answered the way that they did. This can be based on how
they answered other questions as well.

For example, you might believe that website visitors who had
trouble navigating within your website are likely not return
again. If 30% of the respondents said they had trouble
navigating through the website and 40% said they would not
return, you could look at only those that would not return to
determine if poor navigation might be the case. After
filtering to only those who would not return, if 30% or less
said they had trouble navigating, then this is clearly not the
"reason" visitors will not return. By filtering out those that
would return, we expect the percentage to increase
dramatically. If it does, we still cannot conclude that
navigation is "the" reason, only that it might contribute to the
respondents not returning. In order to know if it is "the"
reason, we would need to ask a direct question.
Reporting

After analyzing your survey data, it is time to create a report

of your findings. The complexity and detail need to support
you conclusions, along with your intended audience, will
dictate the format of your report. CEO's require a different
level of detail than line managers, so for maximum results
consider who is going to receive your report and tailor it to
meet their unique needs.

Visual reports, such as an HTML document or Microsoft

PowerPoint presentation, are best suited for simple findings.
These graphical reports are best when they are light on text
and heavy on graphs and charts. They are reviewed quickly
rather than studied at length, and most conclusions are
obvious, so detailed explanations are seldom required. For
more complex topics, a detailed report created in Microsoft
Word or Adobe Acrobat is often required. Reports created
using Word often include much more detailed information,
report findings that require significant explanation, are
extremely text heavy, and are often studied at great length
and in significant detail.

No matter which type of report you use, always remember

that information can be more powerfully displayed in a
graphic format verses a text or tabular representation. Often,
trends and patterns are more obvious and recommendations
more effective when presented visually. Ideally, when
making comparisons one or more groups of respondents, it is
best to show a chart of each group's responses side-by-side.
This side-by-side comparison allows your audience to
quickly see the differences you are highlighting and will lead
to more support for your conclusions.

At the beginning of your report, you should review your

survey objective and sampling method. This will help your
audience understand what the survey was about, and enable
you to avoid many questions that are outside of your original
objectives. Your report should have a description of your
sampling method, including who was invited to participate,
over what time frame results were collected, and any issues
that might exist relative to your respondent pool. Next, you
should include your analysis and conclusions in adequate
detail to meet the needs of your audience. Include a table or
graph for each area of interest and explain why it is
noteworthy. After your analysis section, you should make
recommendations that relate back to your survey objectives.
Recommendations can be as simple as conduct further
studies to a major shift in company direction. In either case,
your recommendation must be within the scope of your
survey objective and supported by the data collected. Finally,
you can include a copy of your survey questions and a
summary of all the data collected as an appendix to your
report.
Conclusion

Survey analysis is not as easy as downloading results and

printing a chart or report, yet it is not so complex that it
requires a PhD. In this article we have learned that good
analysis begins with good questions, representative
participation, and careful interpretation of the data, in order
to produce actionable results. Techniques such as charting,
filtering, cross tabulation, and regression analysis all help
you spot trends and patterns within your data while helping
you meet your survey objective. You now have a solid
foundation upon which you can confidently conduct your
own survey analysis using a tool like eSurveysPro.
This is the html version of the file
http://www.epidemiolog.net/evolving/DataAnalysis-and-
interpretation.pdf.
Google automatically generates html versions of documents
as we crawl the web.
Page 1
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 451
rev. 3/29/2004, 6/27/2004, 7/22/2004
14. Data analysis and interpretation
Concepts and techniques for managing, editing, analyzing
and interpreting data from
epidemiologic studies.
Key concepts/expectations
This chapter contains a great deal of material and goes
beyond what you are expected to learn for
this course (i.e., for examination questions). However,
statistical issues pervade epidemiologic
studies, and you may find some of the material that follows
of use as you read the literature. So if
you find that you are getting lost and begin to wonder what
points you are expected to learn, please
refer to the following list of concepts we expect you to know:
Need to edit data before serious analysis and to catch errors
as soon as possible.
Options for data cleaning – range checks, consistency checks
– and what these can (and can
not) accomplish.
What is meant by data coding and why is it carried out.
Basic meaning of various terms used to characterize the
mathematical attributes of different
kinds of variables, i.e., nominal, dichotomous, categorical,
ordinal, measurement, count,
discrete, interval, ratio, continuous. Be able to recognize
examples of different kinds of
variables and advantages/disadvantages of treating them in
different ways.
What is meant by a “derived” variable and different types of
derived variables.
Objectives of statistical hypothesis tests (“significance”
tests), the meaning of the outcomes
from such tests, and how to interpret a p-value.
What is a confidence interval and how it can be interpreted.
Concepts of Type I error, Type II error, significance level,
confidence level, statistical
“power”, statistical precision, and the relationship among
these concepts and sample size.
Computation of p-values, confidence intervals, power, or
sample size will not be asked for on
exams. Fisher’s exact test, asymptotic tests, z-tables, 1-sided
vs. 2-sided tests, intracluster correlation,
Bayesian versus frequentist approaches, meta-analysis, and
interpretation of multiple significance
tests are all purely for your edification and enjoyment, as far
as EPID 168 is concerned, not for
examinations. In general, I encourage a nondogmatic
approach to statistics (caveat: I am not a
“licensed” statistician!).
Page 2
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 452
rev. 3/29/2004, 6/27/2004, 7/22/2004
Data analysis and interpretation
Epidemiologists often find data analysis the most enjoyable
part of carrying out an epidemiologic
study, since after all of the hard work and waiting they get
the chance to find out the answers. If the
data do not provide answers, that presents yet another
opportunity for creativity! So analyzing the
data and interpreting the results are the “reward” for the work
of collecting the data.
Data do not, however, “speak for themselves”. They reveal
what the analyst can detect. So when
the new investigator, attempting to collect this reward, finds
him/herself alone with the dataset and
no idea how to proceed, the feeling may be one more of
anxiety than of eager anticipation. As with
most other aspects of a study, analysis and interpretation of
the study should relate to the study
objectives and research questions. One often-helpful strategy
is to begin by imagining or even
outlining the manuscript(s) to be written from the data.
The usual analysis approach is to begin with descriptive
analyses, to explore and gain a “feel” for the
data. The analyst then turns to address specific questions
from the study aims or hypotheses, from
findings and questions from studies reported in the literature,
and from patterns suggested by the
descriptive analyses. Before analysis begins in earnest,
though, a considerable amount of preparatory
work must usually be carried out.
Analysis - major objectives
1. Evaluate and enhance data quality
2. Describe the study population and its relationship to some
presumed source (account for all
in-scope potential subjects; compare the available study
population with the target
population)
3. Assess potential for bias (e.g., nonresponse, refusal, and
attrition, comparison groups)
4. Estimate measures of frequency and extent (prevalence,
incidence, means, medians)
5. Estimate measures of strength of association or effect
6. Assess the degree of uncertainty from random noise
(“chance”)
7. Control and examine effects of other relevant factors
8. Seek further insight into the relationships observed or not
observed
9. Evaluate impact or importance
Preparatory work – Data editing
In a well-executed study, the data collection plan, including
procedures, instruments, and forms, is
designed and pretested to maximize accuracy. All data
collection activities are monitored to ensure
adherence to the data collection protocol and to prompt
actions to minimize and resolve missing
Page 3
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 453
rev. 3/29/2004, 6/27/2004, 7/22/2004
and questionable data. Monitoring procedures are instituted
at the outset and maintained
throughout the study, since the faster irregularities can be
detected, the greater the likelihood that
they can be resolved in a satisfactory manner and the sooner
preventive measures can be instituted.
Nevertheless, there is often the need to “edit” data, both
before and after they are computerized.
The first step is “manual” or “visual editing”. Before forms
are keyed (unless the data are entered
into the computer at the time of collection, e.g., through
CATI – computer-assisted telephone
interviewing) the forms are reviewed to spot irregularities
and problems that escaped notice or
correction during monitoring.
Open-ended questions, if there are any, usually need to be
coded. Codes for keying may also be
needed for closed-end questions unless the response choices
are “precoded” (i.e., have numbers or
letters corresponding to each response choice). Even forms
with only closed-end questions having
precoded responses choices may require coding for such
situations as unclear or ambiguous
responses, multiple responses to a single item, written
comments from the participant or data
collector, and other situations that arise. (Coding will be
discussed in greater detail below.) It is
possible to detect data problems (e.g., inconsistent or out of
range responses) at this stage, but these
are often more systematically handled at or following the
time of computerization. Visual editing
also provides the opportunity to get a sense for how well the
forms were filled out and how often
certain types of problems have arisen.
Data forms will usually then be keyed, typically into a
personal computer or computer terminal for
which a programmer has designed data entry screens that
match the layout of the questionnaire. For
small questionnaires and data forms, however, data can be
keyed directly into a spreadsheet or even
a plain text file. A customized data entry program often
checks each value as it is entered, in order to
prevent illegal values from entering the dataset. This facility
serves to reduce keying errors, but will
also detect illegal responses on the form that slipped through
the visual edits. Of course, there must
be some procedure to handle these situations.
Since most epidemiologic studies collect large amounts of
data, monitoring, visual editing, data
entry, and subsequent data checks are typically carried out by
multiple people, often with different
levels of skill, experience, and authority, over an extended
period and in multiple locations. The data
processing procedures need to take these differences into
account, so that when problems are
detected or questions arise an efficient routing is available
for their resolution and that analysis staff
and/or investigators have ways of learning the information
that is gained through the various steps
of the editing process. Techniques such as “batching”, where
forms and other materials are divided
into sets (e.g., 50 forms), counted, possibly summed over one
or two numeric fields, and tracked as a
group, may be helpful to avoid loss of data forms. Quality
control and security are always critical
issues. Their achievement becomes increasingly complex as
staff size and diversity of experience
increase.
Preparatory work – Data cleaning
Once the data are computerized and verified (key-verified by
double-keying or sight-verified) they
are subjected to a series of computer checks to “clean” them.
Page 4
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 454
rev. 3/29/2004, 6/27/2004, 7/22/2004
Range checks
Range checks compare each data item to the set of usual and
permissible values for that variable.
Range checks are used to:
1. Detect and correct invalid values
2. Note and investigate unusual values
3. Note outliers (even if correct their presence may have a
bearing on which statistical methods
to use)
4. Check reasonableness of distributions and also note their
form, since that will also affect
choice of statistical procedures
Consistency checks
Consistency checks examine each pair (occasionally more)
of related data items in relation to the set
of usual and permissible values for the variables as a pair.
For example, males should not have had a
hysterectomy. College students are generally at least 18 years
of age (though exceptions can occur,
so this consistency check is “soft”, not “hard”). Consistency
checks are used to:
1. Detect and correct impermissible combinations
2. Note and investigate unusual combinations
3. Check consistency of denominators and “missing” and
“not applicable” values (i.e., verify
that skip patterns have been followed)
4. Check reasonableness of joint distributions (e.g., in
scatterplots)
In situations where there are a lot of inconsistent responses,
the approach used to handle
inconsistency can have a noticeable impact on estimates and
can alter comparisons across groups.
Authors should describe the decision rules used to deal with
inconsistency and how the procedures
affect the results (Bauer and Johnson, 2000).
Preparatory work – Data coding
Data coding means translating information into values
suitable for computer entry and statistical
analysis. All types of data (e.g., medical records,
questionnaires, laboratory tests) must be coded,
though in some cases the coding has been worked out in
advance. The objective is to create
variables from information, with an eye towards their
analysis. The following questions underlie
coding decisions:
1. What information exists?
2. What information is relevant?
3. How is it likely to be analyzed?
Page 5
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 455
rev. 3/29/2004, 6/27/2004, 7/22/2004
Examples of coding and editing decisions
A typical criterion for HIV seropositivity is a repeatedly-
positive ELISA (enzyme linked
immunosorbent assay) for HIV antibody confirmed with a
Western blot to identify the
presence of particular proteins (e.g., p24, gp41, gp120/160).
Thus, the data from the
laboratory may include all of the following:
a. An overall assessment of HIV status
(positive/negative/indeterminant)
b. Pairs of ELISA results expressed as:
i. + + / + – / – – / indeterminate
ii. optical densities
c. Western Blot results (for persons with positive ELISA
results) expressed as:
i. (+ / – / indeterminant)
ii. specific protein bands detected, e.g., p24, gp41, gp120/160
How much of this information should be coded and keyed?
How to code open-ended questionnaire items (e.g., “In what
ways have you changed your
smoking behavior?”, “What are your reasons for quitting
smoking?”, “What barriers to
changing do you anticipate?”, “What did you do in your
job?”)
Closed-end questions may be “self-coding” (i.e., the code to
be keyed is listed next to each
response choice), but there can also be:
a. Multiple responses where only a single response is wanted
– may be
1. Inconsistent responses (e.g., “Never” and “2 times or
more”)
2. Adjacent responses indicating a range (e.g., “two or three
times” and “four or five
times”, by a respondent who could not choose among 2-5
times).
b. Skipped responses – should differentiate among
1. Question was not applicable for this respondent (e.g., age
at menarche for male
respondents)
2. Respondent declined to answer (which respondents
sometimes may indicate as
“N/A”!)
3. Respondent did not know or could not remember
4. Respondent skipped without apparent reason
It is necessary to achieve a balance between coding the
minimum and coding “everything”.
Coding is much easier when done all at once.
One can always subsequently ignore coded distinctions not
judged as meaningful.
Page 6
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 456
rev. 3/29/2004, 6/27/2004, 7/22/2004
Information not coded will be unavailable for analysis (e.g.,
date questionnaire received,
which questionnaires were randomly selected for 10%
verification survey).
More detail means more recodes for analysis means more
programming means more
opportunities for error.
Decisions deferred have to be made sometime, so why not
decide up front (e.g., When a
respondent circles adjacent response choices, such as “3.
Once or twice” and “4. Two to five
times”, what should be coded – 3?, 4?, 3.5? a missing value
code? a code to be replaced at a
later date when a decision is made?)
It is important to document how coding was done and how
issues were resolved, so that consistency
can be achieved and the inevitable questions (“How did we
deal with that situation?”) answered.
Types of variables - levels or scales of measurement
Constructs or factors being studied are represented by
“variables”. Variables (also sometimes called
“factors”) have “values” or “levels”. Variables summarize
and reduce data, attempting to represent
the “essential” information.
Analytic techniques depend upon variable types
Variables can be classified in various ways. A continuous
variable takes on all values within its
permissible range, so that for any two allowable values there
are other allowable values in between.
A continuous variable (sometimes called a “measurement
variable”) can be used in answer to the
question “how much”. Measurements such as weight, height,
and blood pressure can, in principle,
be represented by continuous variables and are frequently
treated as such in statistical analysis. In
practice, of course, the instruments used to measure these
and other phenomena and the precision
with which values are recorded allow only a finite number of
values, but these can be regarded as
points on a continuum. Mathematically, a discrete variable
can take only certain values between its
maximum and minimum values, even if there is no limit to
the number of such values (e.g., the set
of all rational numbers is countable though unlimited in
number). Discrete variables that can take
any of a large number of values are often treated as if they
were continuous. If the values of a
variable can be placed in order, then whether the analyst
elects to treat it as discrete and/or
continuous depends on the variable’s distribution, the
requirements of available analytic procedures,
and the analyst’s judgment about interpretability.
Types of discrete variables
1. Identification – a variable that simply names each
observation (e.g., a study identifying
number) and which is not used in statistical analysis;
2. Nominal – a categorization or classification, with no
inherent ordering; the values or the
variable are completely arbitrary and could be replaced by
any others without affecting the
results (e.g., ABO blood group, clinic number, ethnicity).
Nominal variables can be
dichotomous (two categories, e.g., gender) or polytomous
(more than two categories).
Page 7
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 457
rev. 3/29/2004, 6/27/2004, 7/22/2004
3. Ordinal – a classification in which values can be ordered
or ranked; since the coded values
need only reflect the ranking they can be replaced by any
others with the same relative
ranking (e.g., 1,2,5; 6,22,69; 3.5,4.2, 6.9 could all be used in
place of 1,2,3). Examples are
injury severity and socioeconomic status.
4. Count – the number of entities, events, or some other
countable phenomenon, for which
the question “how many” is relevant (e.g., parity, number of
siblings); to substitute other
numbers for the variable’s value would change its meaning.
In epidemiologic data analysis,
count variables are often treated as continuous, especially if
the range is large.
Types of continuous variables
1. Interval – differences (intervals) between values are
meaningful, but ratios of values are not.
That is, if the variable takes on the values 11-88, with a mean
of 40, it is meaningful to state
that subject A’s score of 60 is “twice as far from the mean”
as subject B’s score of 50. But it
is not meaningful to say that subject A’s score is “1.5 times
the mean”. The reason is that
the zero point for the scale is arbitrary, so values of the
scores have meaning only in relation
to each other. Without loss of information, the scale can be
shifted: 11-88 could be
translated into 0-77 by subtracting 11. Scale scores can also
be multiplied by a constant.
After either transformation, subject A’s score is still twice as
far from the mean as is
subject B’s, but subject A’s score is no longer 1.5 times the
mean score. Psychological scales
(e.g., anxiety, depression) often have this level of
measurement. An example from physics is
temperature measured on the Fahrenheit or Celsius scale.
2. Ratio – both differences and ratios are meaningful. There
is a non-arbitrary zero point, so it
is meaningful to characterize a value as “x” times the mean
value. Any transformation other
than multiplying by a constant (e.g., a change of units) will
distort the relationships of the
values of a variable measured on the ratio scale.
Physiological parameters such as blood
pressure or cholesterol are ratio measures. Kelvin or absolute
temperature is a ratio scale
measure.
Many variables of importance in epidemiology are
dichotomous (i.e., nominal with two levels) – case
vs. noncase, exposed vs. unexposed. For an apparently
ordinal or continuous variable, the
phenomenon itself may not warrant treatment as such. It is
necessary to ask such question as: “Is
‘more’ really more?” and “Are thresholds or discontinuities
involved?” Again, the underlying reality
(or, rather, our conceptual model of it) determines the
approach to quantification. Variable values
are often collapsed into a small number of categories for
some analyses and used in their original
form for others.
Preparatory work – Data reduction
Data reduction seeks to reduce the number of variables for
analysis by combining single variables
into compound variables that better quantify the construct.
Variables created during coding attempt
to faithfully reflect the original data (e.g., height, weight).
Often these variables can be used directly
for analysis, but it is also often necessary to create additional
variables to represent constructs of
interest. For example, the construct overweight is often
represented by a variable derived from the
values for height and weight. Data reduction includes
simplifying individual variables (e.g., collapsing
Page 8
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 458
rev. 3/29/2004, 6/27/2004, 7/22/2004
six possible values to a smaller number) and deriving
compound variables (e.g. “socioeconomic
status” derived from education and occupation).
In general:
Simpler is better
Avoid extraneous detail
Create additional variables, rather than destroy the original
ones (never overwrite the raw
data!).
Inspect detail before relying on summaries
Verify accuracy of derived variables and recodes by
examining crosstabulations between the
original and derived variables.
Take into account threshold effects, saturation phenomena,
and other nonlinearities
Categorize based on the nature of the phenomenon (e.g., a
study of Down’s syndrome can
collapse all age categories below 30 years; a study of
pregnancy rates will require a finer
breakdown below 30 years and even below 20 years).
Types of derived variables
Scales - In a pure scale (e.g., depression, self-esteem) all of
the items are intended as individual
measures of the same construct. The scale score is usually
the sum of the response values for the
items, though items with negative valence (e.g., “I feel
happy” in a depression scale) must be
inverted. The purpose of deriving a scale score by having
multiple items is to obtain a more reliable
measure of the construct than is possible from a single item.
Scale reliability (internal consistency) is
typically assessed by using Cronbach’s coefficient alpha,
which can be thought of as the average of
all of the inter-item correlations. If the items did indeed
measure the same construct in the same
way and were indeed answered in an identical manner, then
the only differences in their values
should be due to random errors of measurement. Cronbach’s
alpha gives the proportion of the total
variation of the scale scores that is not attributable to random
error. Values of 0.80 or greater are
considered adequate for a scale that will be used to analyze
associations (if the scale is used as a
clinical instrument for individual patients, its alpha should be
at least 0.90 – see Nunally’s textbook,
Psychometrics). When the scale consists of separate
subscales, internal consistency may be more
relevant for the individual subscales than for the scale as a
whole. Analyses of relationships between
individual items (inter-item correlation or agreement),
between each item and the remaining items
(item-remainder correlation), between each item and the total
scale (item-scale correlation), and
among groups of items (factor analysis) are standard methods
of analyzing item performance.
Indexes - An index consists of a group of items that are
combined (usually summed) to give a
measure of a multidimensional construct. Here, each of the
items measures a different aspect or
dimension, so that internal consistency measures like
Cronbach’s alpha are either not relevant or
require a different interpretation. Examples of indexes
derived from several variables include
Page 9
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 459
rev. 3/29/2004, 6/27/2004, 7/22/2004
socioeconomic status (e.g., occupation, income, education,
neighborhood), social support (e.g.,
marital status, number of close family members, number of
close friends), sexual risk behavior
(number of partners, types of partners, use of condoms, anal
intercourse). Items may have different
weights, depending upon their relative importance and the
scale on which they were measured.
Algorithms - A procedure that uses a set of criteria according
to specific rules or considerations,
e.g., major depressive disorder, “effective” contraception (I
have not seen this term used to
designate a type of variable before, but I am not aware of any
other term for this concept).
Preparatory work – Exploring the data
Try to get a “feel” for the data – inspect the distribution of
each variable. Examine bivariate
scatterplots and cross classifications. Do the patterns make
sense? Are they believable?
Observe shape – symmetry vs. skewness, discontinuities
Select summary statistics appropriate to the distribution and
variable type (nominal, ordinal,
measurement)
Location - mean, median, percentage above a cut-point
Dispersion - standard deviation, quantiles
Look for relationships in data
Look within important subgroups
Note proportion of missing values
Preparatory work – Missing values
Missing data are a nuisance and can be a problem. For one,
missing responses mean that the
denominators for many analyses differ, which can be
confusing and tiresome to explain. Also,
analyses that involve multiple variables (e.g., coefficient
alpha, crosstabulations, regression models)
generally exclude an entire observation if it is missing a
value for any variable in the analysis (this
method is called listwise deletion). Thus, an analysis
involving 10 variables, even if each has only
5% missing values, could result in excluding as much as 50%
of the dataset (if there is no overlap
among the missing responses)! Moreover, unless data are
missing completely at random (MCAR
– equivalent to a pattern of missing data that would result
from deleting data values throughout the
dataset without any pattern or predilection whatever), then an
analysis that makes no adjustment for
the missing data will be biased, because certain subgroups
will be underrepresented in the available
data (a form of selection bias).
Imputation for missing values - optional topic
As theory, methods, and computing power have developed
over the years, analytic methods
for handling missing data to minimize their detrimental
effects have improved. These
Page 10
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 460
rev. 3/29/2004, 6/27/2004, 7/22/2004
methods seek to impute values for the missing item
responses in ways that attempt to
increase statistical efficiency (by avoiding the loss of
observations which have one or a few
missing values) and reduce bias. Earlier methods of
imputation, now out of favor, include
replacing each missing value by the mean or median for that
variable. Even though such
practices enable all observations to be used in regression
analyses, these methods do not
reduce bias and tend to introduce additional distortion. More
sophisticated methods reduce
bias from missing data while minimizing distortion from
imputation. These methods derive
imputations that make use of the values of variables for
which data are present and which
are related to the variable being imputed.
Typically, complete data cases (observations that have no
missing values for the variables
of interest) serve as the raw material for the imputations.
Factors that are theoretically
related to the variables to be imputed and with which they
are associated in the complete
data cases are used to develop “predictive” models for the
imputed variables. These models
are then applied to the remaining observations, providing
predicted (“imputed”) values for
their missing responses. The resulting imputations are said to
be conditioned on the
variables in the model.
For example, suppose the available data show a positive
correlation between blood pressure
and age. By conditioning imputations on age, we impute (on
average) higher blood
pressures to older subjects with missing blood pressure data
and lower blood pressures to
younger subjects missing blood pressure data. This technique
preserves the relationship
between age and blood pressure that exists in the complete
data cases. Moreover, if older
subjects are more likely to be missing a blood pressure
reading, then the conditioning
reduces the bias from analyzing only the complete data cases.
If the process that led to the missing data is uniformly
random except for being positively
related to identifiable factors (e.g., subject’s age), then the
missing data process is called
missing at random (MAR), rather than MCAR. In such a
situation, the overall mean
blood pressure for the complete data cases is biased
downwards (due to underrepresentation
of older subjects), but the overall mean based on imputations
conditioned on age is not.
If predicted values are simply substituted for missing values,
however, then although bias
will be reduced so will standard errors. The reason is that the
imputation models were
created based on (imperfect) associations between the
conditioning variables and the
variables being imputed. In contrast, the predicted values are
directly computed from the
model as if, in our example, blood pressure were completely
determined by age. In effect,
the model functions as a “self-fulfilling prophecy”. To avoid
this problem a source of
random variability is introduced into the imputation process.
For example, rather than
substituting the predicted values themselves for the missing
data, the imputed values may be
sampled from distributions whose means are the predicted
values (e.g., if the estimated mean
for a yes-no response were 0.30 [where 1=”yes” and
0=“no”], then the imputed value would
be generated randomly from a binomial distribution with a
proportion of “successes” of
0.30).
Page 11
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 461
rev. 3/29/2004, 6/27/2004, 7/22/2004
In addition, by using multiple imputations (typically five),
the analyst can adjust the standard
errors to reflect the uncertainty introduced by the imputation
process. Carrying out multiple
imputations means repeating the imputation process to create
multiple versions of the
dataset (one for each imputation), analyzing each dataset
separately, and combining the
results according to certain procedures.
Imputation causes the least distortion when the proportion of
missing data is small, and data
are available for variables that are strongly associated with
the variable being imputed.
Perversely, however, imputation is most needed when the
proportion of missing data is
large. Also, unfortunately, the available data may provide
little guidance about whether the
missing process is MCAR, MAR, or “nonignorable”.
Attention to causes of missing
responses during data collection can be helpful (Heitjan,
1997).
[I would like to thank Michael Berbaum and Ralph Folsom
for their patient explanations of
imputation and for reading over earlier versions of this
section.]
Descriptive analyses
Exploration of the data at some point becomes descriptive
analysis, to examine and then to report
measures of frequency (incidence, prevalence) and extent
(means, survival time), association
(differences and ratios), and impact (attributable fraction,
preventive fraction). These measures will
be computed for important subgroups and probably for the
entire study population.
Standardization or other adjustment procedures may be
needed to take account of differences in age
and other risk factor distributions, follow-up time, etc.
Evaluation of hypotheses
After the descriptive analyses comes evaluation of the study
hypotheses, if the study has identified
any. Here there will be a more formal evaluation of potential
confounding, other forms of bias,
potential alternative explanations for what has been
observed. One aspect of both descriptive
analysis and hypothesis testing, especially of the latter, is the
assessment of the likely influence of
random variability (“chance”) on the data. Much of the field
of statistics has grown up to deal with
this aspect, to which we will now turn.
Evaluating the role of chance - inference
Whether or not we believe, in Albert Einstein’s words, that
“the Lord God doesn’t play dice with
the universe”, there are many events in the world that we
ascribe to “chance”. When we roll a die,
the resulting number is generally unpredictable and does not
(or at least, should not) follow any
evident pattern. Similarly, when we draw five cards from a
freshly-shuffled, unmarked deck, we
know that some outcomes are more or less likely than others
(e.g., a pair is more likely than three of
a kind), but we cannot predict what cards we will draw. The
theories of probability and statistics
were born in the gaming parlors of Monte Carlo and came of
age in the fields of the British
countryside. The computer revolution put their power, for
good or for whatever, into the hands of
any of us who can click a mouse.
Page 12
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 462
rev. 3/29/2004, 6/27/2004, 7/22/2004
The basis for the incorporation of the fruits of the theory of
probability and statistics into medical
and epidemiologic research has been recounted by Austin
Bradford Hill as follows:
“Between the two world wars there was a strong case for
emphasizing to the clinician and other
research workers the importance of not overlooking the
effects of the play of chance upon their
data. Perhaps too often generalities were based upon two
men and a laboratory dog while the
treatment of choice was deduced from a difference between
two bedfuls of patients and might
easily have no true meaning. It was therefore a useful
corrective for statisticians to stress, and to
teach the need for, tests of significance merely to serve as
guides to caution before drawing a
conclusion, before inflating the particular to the general.” (pg
299 in The environment and
disease: association or causation. Proceedings of the Royal
Society of Medicine, 1965: 295-300)
From this innocent and commonsensical beginning,
statistical procedures have (like kudzu?? – just
kidding!) virtually engulfed the thinking of researchers in
many fields. Hill continues:
“I wonder whether the pendulum has not swung too far – not
only with the attentive pupils but
even with the statisticians themselves. To decline to draw
conclusions without standard errors
can surely be just as silly? Fortunately I believe we have not
yet gone so far as our friends in the
USA where, I am told, some editors of journals will return an
article because tests of significance
have not been applied. Yet there are innumerable situations
in which they are totally
unnecessary - because the difference is grotesquely obvious,
because it is negligible, or because,
whether it be formally significant or not, it is too small to be
of any practical importance. What
is worse the glitter of the t table diverts attention from the
inadequacies of the fare. . . .”
He admits that he exaggerates, but he suspects that the over-
reliance on statistical tests weakens “our
capacity to interpret data and to take reasonable decisions
whatever the value of P.” Hill is referring
to tests of significance, which are probably the most common
procedures for assessing the role of
chance, or perhaps more precisely, the amount of numerical
evidence that observed differences
would not readily arise by chance alone.
Illustration of a statistical test
Consider the following data, from the first study to report an
association between adenocarcinoma
of the vagina and maternal use of diethylstilbestrol (DES).
During the 1960’s, a handful of cases of
adenocarcinoma of the vagina were observed in young
women, a highly unusual occurrence.
Investigation into the histories of the affected women
revealed that in most cases the girl’s mother
had taken diethylstilbestrol (DES) while she was carrying the
girl in her uterus. At that time DES
had been prescribed in the belief that it might prevent
premature delivery in women who had lost
pregnancies. In how many patients would this history have to
emerge for it before the investigators
could be confident that it was not a chance observation? This
question is usually answered by
means of a statistical test.
Page 13
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 463
rev. 3/29/2004, 6/27/2004, 7/22/2004
Prenatal exposure to diethylstilbestrol (DES)
among young women with adenocarcinoma of the vagina
Exposed to
diethylstilbesterol?
Yes
No Total
Cases
7
1
8
Controls
0
32
32
Total
8
33
40
Source: Herbst AL, Ulfelder H, Poskanzer DC.
Adenocarcinoma of the vagina. Association of
maternal stilbestrol therapy with tumor appearance in young
women. New Engl J Med 1971;
284:878-881. [From Schlesselman JJ. Case-Control Studies.
New York, Oxford, l982: 54]
All but one of the cases had a positive history for intrauterine
exposure to DES. In contrast, not
one of 32 controls did. The relative risk from this table
cannot be calculated directly, because of the
zero cell, but adding 0.5 to all four cells yields a relative risk
(OR) of 325, a stronger association than
most of us will ever encounter in our data. However, this
study has only eight cases. Could these
results be due to chance?
A statistical test of significance is a device for evaluating the
amount of numerical data on which an
observed pattern is based, to answer a question like, “How
often could such a strong association
arise completely by chance in an infinite number of
analogous experiments with the same number of
subjects and the same proportion of cases (or of exposed)?”
This question is not quite the same as
“How likely is it that chance produced the association in the
table?” nor as “How much of the
association is due to chance?”. But if such a strong
association would arise only very infrequently by
chance alone, then it is reasonable to suppose that at least
some potentially identifiable factor has
contributed to the observed association. That factor could be
bias, of course, rather than the
exposure, but at least it would be something other than
chance. Conversely, it is also possible that
much stronger associations could readily arise by chance and
yet the one we observed might reflect a
causal process. The significance test simply evaluates the
strength of numerical evidence for
discounting chance as a likely sufficient explanation.
In order to conduct a test of significance, we need to
operationalize the concept of “analogous
experiment”. There’s the rub. What kind of experiment is
analogous to an epidemiologic study, all
the more so an observational one? For the above table, the
significance test that would be used is
Fisher’s Exact Test. The analogous experiment (probability
model) here is equivalent to the
following:
Suppose that you have 40 pairs of socks – 7 pairs of red
socks, and 33 pairs of blue socks. You
want to pack 8 pairs of socks in your travel bag, so without
looking you take 8 pairs at random
and put them in your bag. How many red pairs have you
packed for your trip?
Page 14
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 464
rev. 3/29/2004, 6/27/2004, 7/22/2004
When this “analogous experiment” is repeated a sufficient
number of times, the proportion of trials
in which the bag has 7 red pairs will provide the probability
that chance alone would produce a
situation in which you had packed 7 pairs of red socks. This
probability is the :p-value” for the
significance test of the relationship between adenocarcinoma
of the vagina and maternal DES in the
above table.
Fortunately, the distribution of the number of red pairs in the
bag has already been worked out
theoretically, so that the exact probability can be computed
without having to carry out what in this
case would be a VERY large number of trials. The formula
for the (hypergeometric) distribution is:
n1
n0
Cj
C
(m1 – j)
n1!n0!m1!m0!
= –––––––––––––––––––––––– =
–––––––––––––––––––––––––––
Pr(A=j)
n
n! j! (n1 – j)! (m1 – j)! (n0 – m1 –j)!
C
m1
where Pr(A=j) is the probability of obtaining j red pairs of
socks in the travel bag and m0, m1, n0, n1,
and n are the row and column totals in the table:
Color
Red
Blue
Total
Travel bag
j
m1 – j
m1
In drawer
n1 – j
n0 – m1 – j
m0
Total
n1
n0
n
Here is how the formula is applied:
Red
(DES)
Blue
Total
Packed (cases)
7
1
8
In drawer (controls)
0
32
32
Total
7
33
40
Page 15
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 465
rev. 3/29/2004, 6/27/2004, 7/22/2004
Possible outcomes
(Colors of pairs of
socks in travel case)
Red
Blue
Probability of
each outcome
0
8
.181
1
7
.389
2
6
.302
3
5
.108
4
4
.019
7! 33! 8! 32!
5
3
.0015
––––––––––––
6
2
.00005
{
40! 5! 2! 3! 30!
7
1
4.3 x 10-7
8
0
0
} p-value
1.0000
Comments on the “red socks” model:
1. A model is a system or structure intended to represent the
essential features of the structure
or system that is the object of study. The above model is a
very simplified representation!
2. The model is derived on the basis of certain constraints or
assumptions (e.g., in this case, 8
cases, 7 DES-exposed mothers, and 40 subjects in all –
“fixed marginals” – as well as “all
permutations are equally likely”).
3. The model underlying hypothesis testing assumes a
repeatable experiment and an a priori
specification of the “hypothesis” being tested – a “null”
hypothesis [this is embodied in the
model of “equally likely” permutations] and an “alternative
hypothesis” [this deals with what
results we would regard as inconsistent with the null
hypothesis].
4. The above model is tedious to compute for large tables,
though computers have solved that
problem.
Concept of hypothesis testing (tests of significance)
What we really want to know is: “Is the observed association
due to chance?”, or “How likely is it
that the observed association is due to chance?”. This
probability is sometimes referred to as the
“posterior [a posteriori] probability”, the probability that the
hypothesis is true given the observed
results. (The “prior [a priori] probability” that the hypothesis
is true is our belief in the hypothesis
before we have the results in question). The frequentist
school of statistics, from which significance
testing derives, cannot answer this question directly. Instead,
significance tests and p-values attempt
to provide an indirect answer, by reformulating the question
as: “How often would an association
as strong as that observed occur by chance alone?”. The role
of chance is played by a suitable
probability model, chosen to represent the probability
structure of the data and the study design. But
most epidemiologic studies deviate rather markedly from the
probability models on which statistical
Page 16
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 466
rev. 3/29/2004, 6/27/2004, 7/22/2004
tests are based (e.g., see Sander Greenland, Randomization,
statistics, and causal inference), so
although statistical theory is extremely precise, it must be
thoughtfully applied and thoughtfully
interpreted.
A compromise version of the question that underlies a
significance test is “How consistent are the
numerical data with what would be expected ‘by chance’ – as
played by a suitable probability
model”. The probability model is most often one that
assumes no systematic difference between
groups, partly because such models are easier to derive and
also because it is often convenient for
the hypothesis-testing framework. The result of the
significance test is a probability (the p-value),
which provides a quantitative answer to this compromise
question. (Note: The statistical “null
hypothesis” is rarely of interest from a substantive
perspective. A study hypothesis should be stated
in terms of no association only if that is indeed what the
investigator hopes to demonstrate. In fact,
it is quite difficult to demonstrate the absence of association,
since the evidence for no association is
related to the Type II error probability (1 – statistical power)
for the study, which is generally
considerably greater than the significance level – see below).
The p-value itself can be regarded as a descriptive statistic, a
piece of evidence that bears on the
amount of numerical evidence for the association under
study. When a decision is called for,
though, then some method of assigning an action to the result
of the significance test is needed.
Decision-making entails the risk of making errors. Ideally the
loss function (the costs of errors of
various types) is known explicitly. Under broadly applicable
assumptions, though, the theory of
decision-making provides a technique for decision-making
based on the results of a statistical test.
That technique is statistical hypothesis testing.
As noted, the hypothesis to be tested is generally a “null
hypothesis” (usually designated H0). H0 is
the probability model that will play the role of chance (for
example, the red socks model). In the
present context, that model will be based on the premise that
there is no association. If there is
sufficient numerical evidence to lead us to reject H0, then we
will decide that the converse is true,
that there is an association. The converse is designated as the
“alternate hypothesis” (HA). The
decision-making rule is to reject H0, in favor of HA, if the p-
value is sufficiently small, and to
otherwise accept H0.
Since we have a decision between two alternatives (H0 and
HA) we can make two kinds of errors:
Type I error:
Erroneously reject H0 (i.e., conclude, incorrectly, that data
are not consistent
with the model)
Type II error: Erroneously fail to reject H0 (i.e., conclude,
incorrectly, that data are
consistent with the model)
(The originator of these terms must have been more prosaic
than the originators of the terms
“significance”, “power”, “precision”, and “efficiency”)
Traditionally, the Type I error probability
has received more attention and is referred to as the
“significance level” of the test.
Page 17
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 467
rev. 3/29/2004, 6/27/2004, 7/22/2004
In a strict decision-making mode, the result of the
significance test is “Reject null hypothesis” or
“Fail to reject null hypothesis”. (Note that “fail to reject the
null hypothesis” is not equivalent to
declaring that the null hypothesis is true.) However, rarely
must a decision be made based on a
single study, so it is preferable to report the calculated p-
value (probability that the assumed
probability model would produce data as extreme or more
so). The p-value gives more information
than the statement “results were significant at the 5% level”,
since it quantifies the degree to which
the data are incompatible with “chance” (as played by the
probability model), allowing the reader to
apply his/her tolerance for a Type I error. Note that the p-
value is not a direct index of the strength
of an association in an epidemiologic sense nor of its
biologic, clinical, or epidemiologic
“significance”. The p-value simply assesses the compatibility
of the observed data with the assumed
probability model that serves to represent H0.
There are many methods for obtaining a p-value or
conducting a test of statistical significance. The
choice depends upon the level of measurement of the
variables (dichotomous, nominal polytomous,
ordinal, continuous), the sampling design from which the
data came, and other factors. The
statistical test illustrated above is an “exact” test (Fisher’s
exact test), since it is based on a model that
considers all possible outcomes and in how many ways each
can occur. In an exact test, the
probability model is readily apparent.
Illustration of an asymptotic test
More commonly-used, because they are much simpler to
compute, are asymptotic tests (e.g., chi-
square, t-test). Asymptotic tests are approximations whose
accuracy improves as the sample size
increases, and the underlying probability model on which
they are based tends to be more abstract.
Typically, asymptotic tests are based on the “normal”
(Gaussian) distribution. Why the Gaussian
distribution? Because it offers a number of analytic
advantages and, most especially, because of the
Central Limit Theorem (“one of the most remarkable
theorems in the whole of mathematics”,
Mood and Graybill, 1963:149). The Central Limit Theorem
holds that if we take large enough
random samples from any distribution with a finite variance,
the means of those samples will have
an approximately Gaussian distribution.
The general form for such a test is (see Rothman, Modern
epidemiology, p. 139 or Kleinbaum, Kupper,
and Morgenstern, Epidemiologic research):
a – E(a)
Z = –––––––––
√var(a)
where “a” is the observed value (e.g., the number of exposed
cases), E(a) is the expected value for
“a” under the null hypothesis (a.k.a. analogous experiment),
and var(a) is the variance of “a” under
the null hypothesis. Thus, Z is the number of standard
deviations by which “a” differs from what
would be expected if there were no association and has an
approximate unit normal distribution. (Z
is occasionally written as Χ. (called “chi”
[pronounced “KAI”], a unit normal distribution is
the same
as the square root of a one-degree-of-freedom chi-square
distribution).
Page 18
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 468
rev. 3/29/2004, 6/27/2004, 7/22/2004
The probability associated with being “Z” standard
deviations away from the mean of a normal
distribution can be computed and is readily available in
statistical tables (see table excerpt below).
The value of a normally-distributed random variable is
usually (i.e., probability 95%) less than two
standard deviations from its mean, so if Z exceeds 1.96 we
say “p<.05”, or more precisely, we take
the value we have calculated for Z, look it up in a table of the
normal distribution and read off the
corresponding p-value.
The table excerpt below shows various probabilities derived
from the unit normal distribution. For
example, the probability associated with a distance of 1.645
standard deviations above the mean is
shown in column B (0.05) and is identical to the probability
associated with a distance of 1.645
standard deviations below the mean (since the normal
distribution is symmetrical). The probability
associated with obtaining a value of z that is either above or
below a distance of 1.645 standard
deviations from the mean is shown in column D (0.10). So if
using the formula above (or one of
those below) we obtain a value of Z equal to 1.645, then the
p-value is either 0.05 or 0.10, depending
upon the alternative hypothesis.
Excerpt from a table of the Normal Distribution
z
h
A
B
C
D
E
0.00
0.3989 0.0000 0.5000 0.0000 1.0000 0.5000
0.01
0.3989 0.0040 0.4960 0.0080 0.9920 0.5040
0.02
0.3989 0.0080 0.4920 0.0160 0.9840 0.5080
...
...
...
...
...
...
...
0.8416 0.2800
0.30
0.20
0.60
0.40
0.80
...
...
...
...
...
...
...
1.282 0.1755
0.40
0.10
0.80
0.20
0.90
...
...
...
...
...
...
...
1.645 0.1031
0.45
0.05
0.90
0.10
0.95
...
...
...
...
...
...
...
1.960 0.0585 0.475
0.025
0.95
0.05
0.975
...
...
...
...
...
...
...
2.576 0.0145 0.495
0.005
0.99
0.01
0.995
...
...
...
...
...
...
...
3.090 0.0034 0.499
0.001
0.998
0.002
0.999
...
...
...
...
...
...
...
Legend:
z = number of standard deviations to the right of the mean
h = height of the normal curve for that number of standard
deviations from the
mean
A = area between the mean and z
B = area to the right of z (or to the left of -z)
Page 19
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 469
rev. 3/29/2004, 6/27/2004, 7/22/2004
C = area between -z and +z
D = area beyond |z| (i.e., to the left of -z and the right of +z)
E = area to the left of z
(Source: National Bureau of Standards – Applied
Mathematics Series–23, U.S.
Government Printing Office, Washington, D.C., 1953, as
abstracted in Table A-4 in
Richard D. Remington and M. Anthony Schork, Statistics
with applications to the
biological and health sciences. Englewood Cliffs, NY, 1970.]
One-sided versus two-sided p-values
Recall that the p-value is the probability of obtaining an
association as strong as (or stronger than)
the association that was observed. It turns out, though, that
the phrase “as strong as (or stronger
than)” is ambiguous, because it does not specify whether or
not it is intended to include inverse
associations, i.e., associations in the opposite direction from
the putative association that motivated
the study. For example, if we observe a 2.5 relative risk, does
“as strong” mean only relative risks of
2.5 or larger, or does it also mean relative risks of 0.4 or
smaller? If the former (only 2.5 and larger),
then the corresponding p-value is “one-sided” (or “one-
tailed”). In contrast, if HA is “either greater
than or equal to 2.5 or [inclusive] less than or equal to 0.4”,
then a two-sided p-value is indicated.
[Only one-sided p-values can be interpreted as the
“probability of observing an association as strong
or stronger under the chance model” (Rothman and
Greenland,185).]
The issue of one-sided versus two-sided p-values can arouse
strong emotions. For a given calculated
value of Z, the one-sided p-value is exactly half of the two-
sided p-value. Proponents of two-sided
p-values argue that a one-sided p-value provides an inflated
measure of the statistical significance
(low probability of obtaining results by chance) of an
association. Appropriate situations for using
one-sided p-values are sometimes characterized as ones
where the investigator has no interest in
finding an association in the opposite direction and would
ignore it even it occurred. However, a
posting on the EPIDEMIOL-L listserv asking for situations
of this sort produced very few
persuasive examples.
Here is a dramatical presentation of some of the issues in
choosing between 1-sided and 2-sided
p-values:
The wife of a good friend of yours has tragically died from
lung cancer. Although she was a
life-long nonsmoker, your friend used to smoke quite
heavily. Before her death she had become
an anti-smoking activist, and her last wishes were that your
friend bring suit against R.J. Morris,
Inc., the manufacturer of the cigarette brand your friend used
to smoke. Knowing that he
cannot afford expert consultation, your friend turns to you
and prevails upon you to assist him
with the lawsuit.
In preparation for the trial, the judge reviews with both sides
the standard of evidence for this
civil proceeding. She explains that for the court to find for
the plaintiff (your side) it must
Page 20
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 470
rev. 3/29/2004, 6/27/2004, 7/22/2004
conclude that the association is supported by “a
preponderance of the evidence”, which she
characterizes as “equivalent to a 90% probability that R.J.
Morris’ cigarettes caused the disease”.
The R.J. Morris attorney objects, declaring that, first of all,
only the probability that cigarettes
can cause disease can be estimated, not the probability that
cigarettes did cause the disease. As
the judge is about to say that the judicial interpretation of
probability permits such a conclusion,
the R.J. Morris attorney raises her second objection: since the
plaintiff is basing his case on
scientific evidence, the plaintiff’s case should be held to the
conventional standard of evidence in
science, which requires a significance level of 5%. [Recall
that the significance level is the
probability of a Type I error, which in this case would mean
finding the company responsible
even though your friend’s lung cancer was really due to
chance. If the court were to fail to find
the tobacco company responsible, even though the
company’s cigarettes did cause the cancer,
that would be or a Type II error.]
Seeing an opportunity, you pass a note to your friend, who
passes it on to his attorney. Upon
reading it, his attorney says to the judge “Your Honor, my
client is prepared to accept the
R.J. Morris’ insistence on a 5% significance level, provided
that it is based on a one-sided
alternative hypothesis.” Beginning to regret that she
introduced the probability metaphor, the
judge turns to the R.J. Morris attorney, who is now hastily
conferring with her biostatistician.
After a quick consultation, the R.J. Morris attorney charges
indignantly that plaintiff’s attorney is
trying, through deception, to obtain a lower standard of
evidence. A 5% one-tailed significance
level, she charges, is actually a 10% significance level, since
everyone knows that two-tailed tests
are more appropriate. Your friend’s attorney senses that this
charge will be a telling point with
the judge and anxiously looks back to you for advice on how
to respond.
With your coaching, your friend’s attorney replies that a two-
tailed test is warranted only when
the appropriate alternative hypothesis (HA) is two-sided. The
question in this case is whether R.J.
Morris is or is not liable, i.e., whether their cigarettes did or
did not cause the cancer. This
question corresponds to a one-sided HA, i.e., the court can
(1) reject H0 (no causation) in favor
of the alternative that R.J. Morris is liable or (2) fail to reject
H0, if the court finds the evidence
insufficient. “May it please the court,” she continues, “there
is no issue here that the cigarette
smoke could have acted to prevent the cancer from
occurring, so requiring a two-tailed
alternative hypothesis is tantamount to imposing a
significance level of 2.5%, which is closer to
the standard for a criminal, rather than a civil, trial”.
With the benefit of further consultation, the R.J. Morris
attorney “strenuously objects”.
“Plaintiff may see this case as involving a one-sided HA, but
notwithstanding the tobacco
settlement, as far as the R.J. Morris Company is concerned
the relationship between smoking
and cancer has not been proved. Therefore a finding that
cigarette smoking can in fact prevent
cancer is just as relevant as plaintiff’s contention that the
cigarettes were responsible.”
You are naturally outraged by the R.J. Morris lawyer’s
assertion that the relationship between
smoking and cancer is not proved, but you have to put that
aside as your friend’s lawyer asks
you is it not correct that the significance level is simply a
mechanism for deciding how many
standard deviations away from the mean are required to
exclude chance as an explanation.
Usually, people exclude chance when the statistical test
comes out two standard deviations from
the center of a normal distribution (actually, 1.96 standard
deviations, which corresponds to a
Page 21
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 471
rev. 3/29/2004, 6/27/2004, 7/22/2004
two-tailed 5% significance level). If the judge does accept
the one-tailed 5% significance level,
even with a good argument that because the appropriate HA
is one-sided so that the Type I error
probability is really only 5%, a decision that meets the test of
being only 1.65 standard deviations
from the mean (corresponding to a one-tailed 5%
significance level) may be vulnerable on
appeal. Since the scientific evidence is firm, would it be
better to agree to the two-tailed test?
The judge looks at her watch, and you see beads of
perspiration breaking out on your friend’s
attorney’s forehead. Meanwhile you’re trying to sort through
the issues. You’ve only just
received your epidemiology degree, and you aren’t yet sure
that it works. It’s true that an appeals
court might reject the idea of a one-tailed test, since appellate
judges tend to be conservative,
and R.J. Morris will certainly appeal an adverse judgment.
But then a dark thought jumps into
your mind. What if R.J. Morris has concocted evidence that
will somehow make it appear that
your friend is responsible for his wife’s death from lung
cancer? You know that that is
outlandish, but what if they could? With a two-sided HA, the
court could reject H0 and find your
friend responsible, thereby destroying him financially and
emotionally. “One-sided!”, you cry out
… and then suddenly you wake with a start. The professor
and your fellow students are looking
at you with puzzlement, wondering what question you
thought that you were responding to. As
you emerge from your daydream you hope that you have not
slept through too much of the
lesson and vow to go to bed earlier in the future.
Significance testing in a two-by-two table
For a two-by-two table, the formula can be more easily
expressed for computational purposes by
defining “a” as the contents of a single cell in the table,
conventionally the “a” (upper left corner)
cell, so that E(a) is the value expected for “a” under the null
hypothesis (n1m1/n), and Var(a) is the
variance of “a” under the null hypothesis {(n1n0m1m0)/
[n2(n-1)], based on the hypergeometric
distribution. The test statistic is then simply:
a – n1m1/n
Z = –––––––––––––––––––––––
√{ (n1 n0m1 m0) /[ n2 (n – 1)]}
An equivalent, but more easily remembered computation
formula, is:
(ad – bc)2 (n – 1)
Z = √ Χ2
=
––––––––––––––
√ n1 n0m1 m0
[Note: you may also see the above formula with n, instead of
(n-1) [e.g., Hennekins and Buring,
p. 251 uses T instead of (n-1)]. The reason is that the above
formula gives a Mantel-Haenszel chi-
square statistic (based on the hypergeometric distribution)
instead of the Pearson chi-square statistic
(based on the normal distribution). For large samples the two
are essentially equivalent. There are
parallel formulas for person-time data.]
Page 22
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 472
rev. 3/29/2004, 6/27/2004, 7/22/2004
Exposed to
diethylstilbesterol?
Yes
No Total
Cases
a
b
m1
Controls
c
d
m0
Total
n1
n0
n
Whatever misgivings we may have about the statistical
model and its application, results with as
small a p-value as that obtained in this study will be very
satisfying to practically any investigator
who obtains them. But to appreciate the dynamics of this
procedure, and the problems of
interpretation that arise in more equivocal circumstances, let
us analyze what underlies a small p-
value.
A small p-value (i.e., low probability that results similar to
those observed would be produced by
“chance” [as played by a given statistical model]) reflects:
a strong observed association (or a large observed difference)
or
a large sample size (roughly speaking).
Therefore, when the p-value is not small, there are two
possibilities (ignoring the possibilities of
systematic error, inappropriate statistical model, etc.):
1. the observed association or difference is not strong.
2. the observed association is of a respectable size but the
study size was too small to judge it
“significant.”
How we interpret a failure to obtain a low p-value depends
upon our judgment of the magnitude of
the observed association and of the statistical power of the
study to detect an important real
difference.
If the p-value is small (e.g., less than five percent [typical],
ten percent [less common], or one percent
[for the demanding or rich in data]), the observed results are
somewhat inconsistent with an
explanation based on chance alone, so we are inclined to
view them as having some origin worth
inquiring about (e.g., systematic influences from the way the
study was designed or conducted,
biological or psychosocial processes related to the factors
under study, etc.). If the observed
difference or association is too small to be scientifically or
clinically significant (as opposed to
statistically significant), we will not care to pursue the matter
regardless of the p-value.
Page 23
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 473
rev. 3/29/2004, 6/27/2004, 7/22/2004
If the p-value is not small (i.e., the results are “not
significant”), was an association observed? If no
association was observed, then the appropriate
characterization of the finding is “no association was
observed” (but see below). If an association was observed,
then we can say “an association was
observed but the data were insufficient to discount chance as
an explanation” [not “there was no
association”!].
If no association was observed, then we also need to ask
what were our chances of detecting a
meaningful association if one exists? If statistical power was
low, then we cannot say much. If
statistical power was high, then we can say the data provide
evidence (assuming, always, that bias is
not present) against the existence of a strong association.
If the observed association is strong enough to be important
if it is not due to chance, then the only
conclusion we can draw is that the data do not provide
sufficient evidence to discount an
explanation of chance alone – this is not equivalent to a
conclusion that “an association was not
observed” [since one was] or that “the observed association
is due to chance” [which no one
knows]. Other characterizations often stated are also
unfortunate:
“The observed association is not significant” [which tends to
impugn it]
“The association did not reach statistical significance”
[which implies that the association should
have been stronger – it may be as strong as it should be but
be based on too few subjects.]
Better to say “an association of ____ was observed, but the
data were too few to discount an
explanation based on chance” or some similar expression.
[Note: Any result can become
“nonsignificant” if we stratify enough.]
An alternate possibility is that the observed association was
too weak to be meaningful even if it had
been associated with a small p-value. Here our conclusion
depends upon the size of the study, i.e.,
its statistical power to detect an association of some
particular magnitude. If the power was low, if
the study’s ability to detect a difference we would regard as
important is low, then there really is not
much we can say or conclude, except that our failure to find
an association could well be due to
chance (i.e., we may well have made a “Type II error”). This
inability is one of the reasons for
discouraging researchers from conducting small studies
except as a pilot study to develop
procedures and instruments. If the power was high, then we
are in a better position to interpret our
results as evidence against the existence of a real association.
Statistical power and sample size
Statistical power refers to the ability to detect an association
of interest in the face of sampling error.
Suppose that there is a true association of a certain type and
degree, but that through the workings
of chance our studies will observe the association to be
weaker or stronger. In order to be
reasonably certain that our study will detect the association,
the study has to be large enough so that
sampling error can be contained.
Page 24
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 474
rev. 3/29/2004, 6/27/2004, 7/22/2004
For example, suppose that we are comparing a group of
Alzheimer’s disease cases to a group of
controls to see whether the cases are different in respect to
presence of a specific gene. Suppose
that this gene is actually present in 20% of cases and in 10%
of the population from which the cases
arose (i.e., the OR in a large, unbiased case-control study
would be 2.25). If we study 20 cases and
20 controls, we may well find 4 cases with the gene and 2
controls with the gene, so that we
correctly estimate the prevalence of the gene in cases and in
the population and the OR.
With such few subjects, however, we could very easily get
only 3 cases with the gene and 3 controls
with the gene, completely failing to detect the difference in
prevalence (OR = 1.0). In fact, we might
even get 4 controls with the gene and only 2 cases with the
gene, so that the gene appear to be
protective (OR = 0.44). Of course, we would not want to
react to a difference or an OR that may
readily be due to chance, so we will test whatever result we
observe to make sure that it is greater
than is expected to occur by chance alone (i.e.,”significant”).
That means that we will discount any
association that we regard as within chance expectation. (Or
recalling our courtroom fantasy, a
“preponderance of the evidence”, not merely suspicion.)
Therefore, in order to detect an association, we must both (1)
observe it in our study and (2) decide
that chance would not likely have created it. Each of these
requirements places a demand on the
size of the study. We need at least some minimum number of
subjects so that (1) we have a
reasonable expectation of observing an association if one
exists (i.e., that we will not make a type II
error), and (2) we will think it unlikely that chance could
produce an association of that size.
Statistical power to detect an OR ≠ 1.0 with a one-tailed
significance test
Distribution of test statistic if true OR =1.0
(H0)
z=–1 z=0 z=1 Type I error prob. (alpha) →
zα→
←zβ
← Type II error probability (beta) z=0 z=1
(HA)
Distribution of test statistic if OR is, e.g., 2.25
This diagram illustrates the overlap between the central
portions of the distributions of a test
statistic (e.g., Z) expected under the null hypothesis (e.g..,
true OR is 1.0) and alternate hypothesis
(e.g., true OR is 2.25). When we obtain the results from the
study we will compute the test statistic
(e.g., Z) and compare it to its distribution under the H0 (the
upper of the two distributions in the
diagram). If the calculated value of Z is smaller
than zα, i.e., it falls to the left of the cutpoint we
have set (defined by the Type I error probability, alpha), then
we will conclude that the data we
observed came from the upper distribution (the one for no
association, true OR=1.0). Even if the
Page 25
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 475
rev. 3/29/2004, 6/27/2004, 7/22/2004
OR we observed was greater than 1.0 (which implies that Z
was greater than 0), because Z was not
greater than our cutpoint we regard the observed OR as a
chance deviation from 1.0. If the unseen
truth is that there really is no association, then our conclusion
is correct. If instead the true OR is
really 2.25, so that the data we observed really came from the
lower distribution, then our conclusion
represents a Type II error. The area to the left of the cutpoint
on the lower distribution represents
the probability of making a Type II error, “beta”. Statistical
power – the probability of detecting a
true difference – is equal to one minus beta (i.e., 1 - beta).
Conversely, if we observe a value of Z to the right of the
cutpoint, we will conclude that the data we
observed did not come from the upper distribution and that
therefore the true OR is greater than
1.0. If we are incorrect – if the association we observed was
in fact simply a chance finding – then
our conclusion represents a Type I error. The area to the right
of the cutpoint on the upper
distribution represents the probability of making a Type I
error, “alpha”.
If we abhor making a Type I error, we can move the cutpoint
to the right, which reduces alpha – but
increases beta. If we prefer to reduce beta, we can move the
cutpoint to the left – but that increases
alpha. What we would really like to do is to reduce both
alpha and beta, by making the distributions
narrower (so that more of the shading is located at the center
of the each distribution, symbolizing
greater precision of estimation). The width of the distribution
is controlled by the sample size.
With a powerful light we can easily distinguish between, for
example, a snake and a stick. But with a
weak light, we cannot be certain what we are seeing. We can
elect to err on one side or the other,
but the only way to reduce our chance of error is to get a
more powerful light.
Commonly used values for alpha and beta are, respectively,
0.05 and 0.20 (power=0.80), for a total
probability of error of 0.25. If the study size is limited due to
the low incidence of the disease, the
low prevalence of the exposure, or the low amount of the
budget, then our study estimates will be
imprecise – the distributions in the above diagram will be
wide. The total error probability will be
below 0.25 only when the lower distribution is farther to the
right, i.e., corresponds to a stronger
association.
In essence, intolerance for error (i.e., small alpha and beta)
and desire to detect weak associations
must be paid for with sample size. In our courtroom
daydream, the better the chance we want of
winning the case against R.J. Morris (our power) and/or the
more R.J. Morris can persuade the
judge to raise the standard of evidence (the significance
level), the higher the price we will have to
pay for our legal representation (more study subjects).] The
Appendix contains a section that
translates these concepts into estimated sample sizes.
Small studies bias
In crude terms, big studies are powerful; small studies are
weak. The concept of “small studies bias”
illustrates the importance of having an understanding of
statistical power when interpreting
epidemiologic studies.
Page 26
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 476
rev. 3/29/2004, 6/27/2004, 7/22/2004
The idea behind small studies bias (Richard Peto, Malcolm
Pike, et al., Br J Cancer 34:585-612, 1976)
is that since small studies are easier to carry out than large
studies, many more are carried out. Small
studies that do not find a “significant” result are often not
published. The journals tend not to be
interested, since as explained above, there is not much
information in a negative study that had low
power. In fact, the investigators may not even write up the
results – why not just conduct another
study. In contrast, large studies are expensive and involve
many investigators. Whatever the results
from a large study, there is more interest on everyone’s part
to publish it.
To the extent that this scenario describes reality, the body of
published studies contains primarily
small studies with “significant” results and large studies with
“significant” and “nonsignificant”
results. However, if there are many small (i.e., easy,
inexpensive) studies going on, then a 5%
probability of making a Type I error translates into a large
number of positive findings and resultant
publications. Thus, many of the small studies that appear in
the literature are reporting Type I errors
rather than real associations.
The following example, based on randomized trials of new
treatments, comes from the article by
Peto, Pike, et al. Assume that there are 100 large and 1,000
small trials of treatments that are not
really different, and 20 large and 200 small trials of
treatments which are really different. The large
trials have statistical power of 95%; the small trials have
statistical power of 25%. The significance
level is 5%, and only trials reporting significant results are
published. These somewhat pessimistic,
but perhaps very realistic, assumptions lead to the following
hypothetical scenario for the number of
treatment trials in progress that will be “statistically
significant” (p<0.05):
True death rate in
Expected number to find
Planned
trial size
Control Treatment
Postulated
# of trials
p>0.05
p<0.05
250
50%
50%
100
95 (TN)*
5 (FP)*
250
50%
33%
20
1 (FN)
19 (TP)
25
50%
50%
1,000
950 (TN)
50 (FP)
25
50%
33%
1,000
150 (FN)
50 (TP)
* TN, FP, FN, TP are for analogy with sensitivity and
specificity (see below).
In this scenario, 100 small trials with “significant” results
will be published, but only half of them
will reflect a real difference between treatments. Peto, Pike et
al.’s conclusion is to pay attention only
to large trials, particularly ones that are large enough to be
published even if they do not find a
significant difference in treatments.
These results can be thought of in terms of the concepts of
sensitivity, specificity, and predictive
value. In this conceptualization, sensitivity corresponds to
the statistical power to detect a true
difference (95% for large trials, 25% for small trials),
specificity corresponds to one minus the
significance level – the probability of correctly identifying a
chance result (95% specificity for a 5%
significance level), and positive predictive value is the
probability that a “significant” result in fact
reflects a true difference in treatment effectiveness.
Page 27
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 477
rev. 3/29/2004, 6/27/2004, 7/22/2004
Large trials (e.g., 250 deaths)
True death rate in
treatment group
(assuming 50% death
rate in control group)
P < 0.05
33%
50%
Total
Yes
19
5
24
No
1
95
96
Total
20
100
120
Thus, the predictive value of a p < 0.05 = 19/24 = 79%
Small trials (e.g., 25 deaths)
True death rate in
treatment group
(assuming 50% death
rate in control group)
P < 0.05
33%
50%
Total
Yes
50
50
100
No
150
950
1,100
Total
200
1,000
1,200
Predictive value of a P<.05 = 50/100 = 50%
Evaluating the role of chance - interval estimation
[EPID 168 students are responsible for these concepts, but
not for the computations]
Statistical significance testing, with its decision-making
orientation, has fallen somewhat out of favor
for reporting data from epidemiologic investigations. On the
premise that an epidemiologic study is
essentially a measurement process (see Rothman), it is
argued that the more appropriate statistical
approach is one of estimation (e.g., of a measure of effect)
rather than significance testing. Of
course, there is still a need to quantify the role of chance, but
in an estimation framework chance is
quanitified by a confidence interval or confidence limits
about the point estimate. Confidence limits
quantify the amount of uncertainty in an estimate by defining
an interval which should cover the
population parameter being estimated (e.g., measure of
effect) a known percentage of the time.
Various authors have argued that confidence intervals are
superior to p-values as a means of
quantifying the degree of random error underlying an
association.
Page 28
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 478
rev. 3/29/2004, 6/27/2004, 7/22/2004
Confidence intervals address the question, “what possible
values for a population parameter (e.g.,
incidence density ratio) are consistent with the observed
results?” Stated another way, “what is the
range of true values which, when distorted by haphazard
influences, could well have produced the
observed results?” Confidence intervals provide a statement
about the precision of an estimate or
estimates based on the amount of data available for the
estimate. If a “significant” association was
not observed, then the confidence interval can give some idea
of how strong an association might
nevertheless exist but, due to the luck of the draw, not be
observed.
The nature of a confidence interval and what it does and does
not provide, however, is a little tricky
(judging from a discussion of confidence intervals on the
STAT-L internet listserv that continued
for weeks and drew a host of responses and counter-
responses). The frequentist view is that a “95%
confidence interval” is an interval obtained from a procedure
that 95% of the time yields an interval
containing the true parameter. Ideally, a 95% confidence
interval would be one that “contains the
parameter with 95% probability”. But frequentists argue that
the interval is set by the data, and the
population parameter already exists in nature. The parameter
is either in the interval or it is not.
There is no probability about it. All that can be said is that
95% of the time the procedure will yield
an interval that embraces the value of the parameter (and
therefore 5% of the time the procedure
will yield an interval that does not). In this view, a 95%
confidence interval is like a student who
typically scores 95% – the probability that he/she will give
the correct answer to a question is 95%,
but the answer he/she gave to any particular question was
either correct or incorrect.
Page 29
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 479
rev. 3/29/2004, 6/27/2004, 7/22/2004
Possible true values
Possible true values
Observed
result
Observed
result
Confidence interval
The concept behind the confidence interval
Page 30
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 480
rev. 3/29/2004, 6/27/2004, 7/22/2004
Computing a confidence interval for a ratio measure of effect
Introductory biostatistics courses cover the method for
obtaining a 95% confidence interval for the
estimate of a population proportion p. If the sample is large
enough so that np > 5 and n(1-p) > 5,
then the confidence limits are:
p ± 1.96 √ [var(p)]
p ± 1.96 √[p(1 – p)/n]
where p is the observed proportion, var(p) is the variance of
the estimate of p (so √var (p) is the
standard error), and n is the number of observations. For a
proportion, var(p) equals p(1 – p)/n.
This method can be used to estimate confidence intervals for
prevalence, cumulative incidence, and
other simple proportions. Many epidemiologic measures,
however, are ratios (e.g., CIR, IDR, and
OR). Since ratio measures of effect have a highly skewed
distribution (most of the possible values
lie to the right of the null value of 1.0), the usual approach is
to first estimate the confidence interval
for the natural logarithm [ln(CIR), ln(IDR), or ln(OR)] and
then take the anti-log (exponent) of the
confidence limits:
95% CI for ln(OR) = ln(OR) ± 1.96 √{var[ln(OR)]}
95% CI for OR = exp{ln(OR) ± 1.96 √([var[ln(OR)])}
= OR exp{± 1.96 √([var([ln(OR)])}
To obtain the variance of the ln(OR), we use a simple
formula (that has been derived by means of a
Taylor series approximation to the ln[OR]):
var{[ln(OR)] = 1/a + 1/b + 1/c + 1/d}
which works well if a, b, c, d are all at least 5.
The 95% confidence interval for the ln(OR) is therefore:
ln(OR) + 1.96 √[(1/a + 1/b + 1/c + 1/d)]
and the 95% confidence interval for the OR is:
OR exp{+ 1.96 √ (1/a + 1/b + 1/c + 1/d)}
or
OR e+ 1.96 √ (1/a + 1/b + 1/c + 1/d)
Page 31
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 481
rev. 3/29/2004, 6/27/2004, 7/22/2004
Formulas for confidence intervals for the CIR and IDR can
be found in Kleinbaum, Kupper and
Morgenstern and Rothman and Greenland. Of course, if the
study population is highly-selected
(i.e., unrepresentative of any other population of interest),
how useful is the value of the estimate?
IMPORTANT CAVEAT: Everything in this section, of
course, has been based on the assumption
of perfect (unbiased, independent) sampling and
measurement. Anything other than an unbiased
simple random sample and any error in measurement will
invalidate the above at least to some
extent.
Meta-analysis
Meta-analysis is a quantitative approach to summarizing and
synthesizing the findings from different
studies of a particular relationship of interest. Meta-analysis
proceeds from the recognition that the
failure to find “significant results” can be due as much to the
limited statistical power of individual
studies as to the absence of a relationship. Combining the
information from multiple studies can
yield a more precise and definitive assessment of the
existence and strength of a relationship than is
available from any one study or, it is argued, from a
nonquantitative distillation of the literature.
There are four steps in carrying out a meta-analysis: 1)
formulating the problem, 2) identifying the
studies (published and unpublished), 3) coding and
evaluating the studies, and 4) statistical analysis.
Steps 2) and 3) are critical for the validity of the meta-
analysis, since the judgments from the meta-
analysis will depend upon the adequacy with which the
evidence about the relationship is
represented by the studies that are finally analyzed (the
possibility of publication bias against
“negative” studies implies that some effort should be made to
locate unpublished studies). The
strategy for statistical analysis can be similar to that for
stratified analysis, regarding each study as a
separate “stratum”. More refined approaches recognize that
the studies themselves can be regarded
as a sample from some universe of possible studies, so that
the weighting scheme needs to take into
account inter-study variability as well as intra-study
variability (as in the random-effects model of
analysis of variance).
In its pure form, meta-analysis is predicated on the
assumption that the collection of studies
represents a random sample of equivalently-obtained
observations of an association, so that the
differences across the studies can be regarded as random
(sampling) variability. Hence, a summary
constructed by combining the studies gives a more precise
estimate of the true association. In actual
practice, however, epidemiologic studies are rarely
equivalent, since they often differ in the
population(s) studied, measurements taken, and analysis
approaches. Even studies that appear to be
equivalent (e.g., “unmatched population-based case-control
studies with a physiologic measure of
exposure and control for the same set of potential
confounders”) will differ in less obvious ways: the
populations likely differ in unknown and unmeasured ways,
the disease ascertainment systems may
differ across populations, response factors in the selection of
controls may differ, collection
processes and laboratory analysis of the exposure may differ
in subtle ways that can nevertheless
affect the results (e.g., see examples involving HIV tests and
homocysteine analyses in J Clin
Epidemiol 2001(5)), and collection of data and analytic
handling of potential confounders can differ.
An exploration of heterogeneity in the meta-analysis of
studies of SIDS and sleeping positions
(Dwyer et al, 2001) illustrates some of these issues.
Page 32
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 482
rev. 3/29/2004, 6/27/2004, 7/22/2004
Interpretation of results
Key questions
1. How good are the data?
2. Could chance or bias explain the results?
3. How do the results compare with those from other studies?
4. What theories or mechanisms might account for findings?
5. What new hypotheses are suggested?
6. What are the next research steps?
7. What are the clinical and policy implications?
Page 33
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 483
rev. 3/29/2004, 6/27/2004, 7/22/2004
Bibliography
General
Ahlbom, Anders. Biostatistics for epidemiologists. Boca
Raton, Florida, Lesis Publishers, 1993, 214 pp.,
$45.00 (reviewed in Am J Epidemiol, April 15, 1994).
Bailar, John C., III; Thomas A. Louis, Philip W. Lavori,
Marcia Polansky. Studies without internal
controls. N Engl J Med 1984; 311:156-62.
Bauer UE, Johnson TM. Editing data: what difference do
consistency checks make. Am J Epidemiol
2000151:921-6.
Bulpitt, C.J. Confidence intervals. The Lancet 28 February
1987: 494-497.
Dwyer, Terence; David Couper, Stephen D. Walter. Sources
of heterogeneity in the meta-analysis
of observational studies: The example of SIDS and sleeping
position. J Chron Dis 2001;54:440-447.
Feinstein, Alvan R. The fragility of an altered proportion: a
simple method for explaining standard
errors. J Chron Dis 1987; 40:189-192.
Feinstein, Alvan R. X and iprr: An improved summary for
scientific communication. J Chron Dis
1987; 40:283-288.
Frank, John W. Causation revisited. J Clin Epidemiol 1988;
41:425-426.
Gerbarg, Zachary B.; Ralph I. Horwitz. Resolving conflicting
clinical trials: guidlines for meta-
analysis. J Clin Epidemiol 1988; 41:503-509.
Glantz, Stanton A. Primer of biostatistics. NY, McGraw-Hill,
1981.
Godfrey, Katherine. Comparing means of several groups. N
Engl J Med 1985;313:1450-6.
Hertz-Picciotto, Irva. What you should have learned about
epidemiologic data analysis.
Epidemiology 1999;10:778-783.
Northridge, Mary E.; Bruce Levin, Manning Feinleib,
Mervyn W. Susser. Statistics in the journal–
significance, confidence, and all that. Editorial. Am J Public
Hlth 1997;87(7):1092-1095.
Powell-Tuck J, MacRae KD, Healy MJR, Lennard-Jones JE,
Parkins RA. A defence of the small
clinical trial: evaluation of three gastroenterological studies.
Br Med J 1986; 292:599-602.
Page 34
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 484
rev. 3/29/2004, 6/27/2004, 7/22/2004
Ragland, David R. Dichotomizing continuous outcome
variables: dependence of the magnitude of
association and statistical power on the cutpoint.
Epidemiology 1992;3:434-440
Rothman - Modern Epidemiology, Chapters 9, 10, 14.
Schlesselman - Case-control studies, Chapters 7-8.
(Especially the first few pages of each of these
chapters).
Woolf SH, Battista RN, Anderson GM, Logan AG, et al.
Assessing the clinical effectiveness of
preventive maneuvers: analytic principles and systematic
methods in reviewing evidence and
developing clinical practice recommendations. J Clin
Epidemiol 1990; 43:891-905.
Zeger, Scott L. Statistical reasoning in epidemiology. Am J
Epidemiol 1991; 134(10):1062-1066.
The role of statistical hypothesis tests, confidence intervals,
and other summary
measures of statistical significance and precision of estimates
Allan H. Smith and Michael N. Bates. Confidence limit
analyses should replace power calculations
in the interpretation of epidemiologic studies. Epidemiology
1992;3:449-452
Browner, Warren S.; Thomas B. Newman. Are all significant
P values created equal? JAMA 1987;
257:2459-2463.
Fleiss, Joseph L. Significance tests have a role in
epidemiologic research: reactions to A.M. Walker
(Am J Public Health 1986; 76:559-560). See also
correspondence (587-588 and 1033).
George A. Diamond and James S. Forrester. Clinical trials
and statistical verdicts: probable grounds
for appeal. Annals of Internal Medicine 1983; 93:385-394
Greenland, Sander. Randomization, statistics, and causal
inference. Epidemiology 1990;1:421-429.
Maclure, Malcome; Greenland, Sander. Tests for trend and
dose response: misinterpretations and
alternatives. Am J Epidemiol 1992;135:96-104.
Mood, Alexander M. and Franklin A. Graybill. Introduction
to the theory of statistics. 2ed. NY, McGraw-
Hill, 1963.
Oakes, Michael. Statistical inference. Chestnut Hill, Mass.,
Epidemiology Resources, 1986.
Peace, Karl E. The alternative hypothesis: one-sided or two-
sided? J Clin Epidemiol 1989; 42(5):473-
477.
Page 35
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 485
rev. 3/29/2004, 6/27/2004, 7/22/2004
Poole, Charles. Beyond the confidence interval Am J Public
Health 1987; 77:195-199.
Poole, C. Confidence intervals exclude nothing Am J Public
Health 1987; 77:492-493. (Additional
correspondence (1987; 77:237)).
Savitz DA, Tolo KA, Poole C. Statistical significance testing
in the American Journal of Epidemiology,
1970-1990. Am J Epidemiol 1994;139:1047-.
Thompson, W. Douglas. Statistical criteria in the
interpretation of epidemiologic data Am J Public
Health 1987; 77:191-194.
Thompson, W.D. On the comparison of effects Am J Public
Health 1987; 77:491-492.
Walker, Alexander M. Reporting the results of epidemiologic
studies Am J Public Health 1986;
76:556-558.
Woolson, Robert F., and Joel C. Kleinman. Perspectives on
statistical significance testing. Annual
Review of Public Health 1989(10).
Sample size estimation
Donner A, Birkett N, and Burk C. Randomization by Cluster:
sample size requirements and
analysis. Am J Epidemiol 1981; 114:706
Snedecor GW, Cochran WG. Statistical Methods, 1980 (7th
ed) see pages 102-105, 129-130 (Table A
is from page 104).
Imputation
Heitjan, Daniel F. Annotation: what can be done about
missing data? Approaches to imputation.
Am J Public Hlth 1987;87(4):548-550.
Little RJA, Rubin DB. Statistical analysis with missing data.
NY, Wiley, 1987.
Interpretation of multiple tests of statistical significance
Bulpitt, Christopher J. Subgroup analysis. Lancet 1988 (July
2);31-34.
Cupples, L. Adrienne; Timothy Heeren, Arthur Schatzkin,
Theodore Coulton. Multiple testing of
hypotheses in comparing two groups. Annals of Internal
Medicine 1984; 100:122-129.
Holford, Theodore R.; Stephen D. Walter, Charles W.
Dunnett. Simultaneous interval estimates of
the odds ratio in studies with two or more comparisons. J
Clin Epidemiol 1989; 42(5):427-434.
Page 36
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 486
rev. 3/29/2004, 6/27/2004, 7/22/2004
Jones, David R. and Lesley Rushton. Simultaneous inference
in epidemiological studies. Int J
Epidemiol 1982;11:276-282.
Lee, Kerry L., Frederick McNeer, Frank Starmer, et al.
Lessons from a simulated randomized trial
in coronary artery disease. Circulation 61:508-515, 1980.
Stallones, Reuel A. The use and abuse of subgroup analysis
in epidemiological research. Preventive
Medicine 1987; 16:183-194 (from Workshop on Guidelines
to the Epidemiology of Weak
Associations)
See also Rothman, Modern Epidemiology.
Interpretation of “negative” studies
Freiman, Jennie A., Thomas C. Chalmers, Harry Smith, Jr.,
and Roy R. Kuebler. The importance of
beta, the Type II error and sample size in the design and
interpretation of the randomized control
trial. N Engl J Med 1978;299:690-694.
Hulka, Barbara S. When is the evidence for ‘no association’
sufficient? Editorial. JAMA 1984;
252:81-82.
Meta-analysis
Light, R.J.; D.B. Pillemer. Summing up: the science of
reviewing research. Cambridge MA, Harvard
University Press, 1984. (very readable)
Longnecker M.P.; J.A. Berlin, M.J. Orza, T.C. Chalmers. A
meta-analysis of alcohol consumption in
relation to risk of breast cancer. JAMA 260(5):652-656.
(example)
Wolf, F.M. Meta-Analysis: quantitative methods for research
synthesis. Beverly Hills, CA, Sage, 1986.
Bias
Greenland, Sander. The effect of misclassification in the
presence of covariates. Am J Epidemiol
1980; 112:564-569.
Walter, Stephen D. Effects of interaction, confounding and
observational error on attributable risk
estimation. Am J Epidemiol 1983;117:598-604.
Page 37
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 487
rev. 3/29/2004, 6/27/2004, 7/22/2004
Appendix
Estimating sample size to compare two proportions or means
(Adapted from a summary provided by of Dana Quade, UNC
Department of Biostatistics, June
1984)
Let N be the number of subjects (observational units)
required in each of two groups to be
compared. Then
N=IxDxC
Where:
I = Intolerance for error, which depends on:
a. Alpha = Desired significance level that we want to use for
our hypothesis test (e.g., 5%,
two-sided)
b. Beta = Type II error (e.g.,.10 – same as 1 - power)
Formula: I = (Zalpha + Zbeta)2
Zalpha and Zbeta are, respectively, the critical values
corresponding to alpha and beta from
the normal distribution (see Table A on next page)
D = Difference to detect, which depends on the narrowness
of the difference between the true
proportions or means, in relation to the standard deviation of
that difference. D can be
regarded as the inverse of the “signal-to-noise ratio” – the
softer the signal or the louder the
noise, the more subjects needed
“noise”
p1 (1 – p1) + p2 (1 – p2)
2(σ2)
D = ––––––– OR ––––––––––––––––––––– OR
––––––––
“signal”
(p1 – p2)2
(μ1 – μ2)2
(for differences in
proportions, where (p1 and
p2 are the two proportions –
see table on next page)
(for differences in
means, where μ1 and
μ2 are the two
means, and σ2 is the
variance of the
difference)
Page 38
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 488
rev. 3/29/2004, 6/27/2004, 7/22/2004
C - Clustered observations, which depends on whether
observations are selected independently
or in clusters.
If all observations are sampled independently, C = 1.
If observations are sampled in clusters (e.g., by households,
schools, worksites, census tracts,
etc.), then sample size must be increased to offset the fact
that observations within a cluster
are more similar to each other than to observations in other
clusters. If rho is the
intracluster correlation among observations within clusters,
then:
C = 1 + (m-1) rho
where m is the average cluster size (i.e., n = km, where k is
the number of clusters). C is
often referred to as the design effect. If the clusters are large
or if the people in them tend
to be very similar, then individual subjects contribute little
information and you therefore
need to study a very large number of them. If you choose
“independent thinkers”, you will
learn more from each one.
Table A: Intolerance for error
Two-Tailed Tests
One-Tailed Tests
Significance Level
Significance Level
Desired
power
0.01
0.05
0.10
0.01
0.05
0.10
0.80
11.7
7.9
6.2
10.0
6.2
4.5
.90
14.9
10.5
8.6
13.0
8.6
6.6
0.95
17.8
13.0
10.8
15.8
10.8
8.6
Page 39
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 489
rev. 3/29/2004, 6/27/2004, 7/22/2004
Table B: Difference to be detected
p2
.10
.20
.30
.40
.50
.60
.05 55
9.2
4.1
2.4
1.5
1.0
.10
–
25
7.5
3.7
2.1
1.3
p1
.15 87 115
15
5.9
3.1
1.8
.20 25
–
37
10
4.6
2.5
.25 12.3
139 159
19
7.0
3.5
Complications
1) Unequal sample sizes
Let n be the average sample size = (n1+n2)/2
Let lambda1 = n1/2n, lambda2 = n2/2n (lambda1 + lambda2
= 1)
p1 (1 – p1) p2 (1 – p2)
σ2
1
σ2
2
––––––––– + –––––––––
––––––––– + –––––––––
2 lambda1
2 lambda2
2 lambda1
2 lambda2
D = –––––––––––––––––––––––––– OR
––––––––––––––––––––––––
(p1 – p2)2
(μ1 – μ2)2
2) Covariables
If statistical tests are to be conducted separately within each
stratum then n as determined above is
required for each stratum.
If results for different strata are to be tested only for an
overall average association, it is probably
best not to try to allow for them in the sample size formulas
explicitly, but make a modest overall
increase in n.
Page 40
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 490
rev. 3/29/2004, 6/27/2004, 7/22/2004
Note: more “precise” formulas can be found in the literature,
but the parameters needed for factors
D and C are never really known.
Sample size for interval estimation
The tolerable width for a confidence interval can be used as
the target for estimating the required
sample size for a study population. Suppose, for example,
that an investigator wishes to estimate the
proportion (p) of condom use in a clinic population. If the
investigator can obtain a simple random
sample of that population, then her estimate of the proportion
of condom users would be p = u/n,
where u is the number of users in the sample and n is the size
of her sample. As noted above, if
np > 5 and n(1-p) > 5, then a 95% confidence interval for p is
:
p + 1.96 √[p(1 – p)/ n]
For example, if p is 0.50, then the confidence interval is:
(0.5)
0.5 + 1.96 √[ (0.5)(0.5)/n]
= 0.5 + 1.96 ––––
√n
[The square root of (0.5)(0.5) is, of course, 0.5]
Since 1.96 × 0.5 is approximately 1, for practical purposes
this expression is equivalent to:
0.5 + 1/√ n , so that the confidence limits are (0.5 – 1/√ n ,
0.5 + 1/√ n )
For example, suppose that n, the sample size, is 100. Then
the 95% confidence interval around the
point estimate of 0.5 is:
(0.5 – 1/√ 100 , 0.5 + 1/√ 100 )
=
(0.5 – 1/10 , 0.5 + 1/10)
=
(0.5 – 0.1, 0.5 + 0.1)
=
(0.4, 0.6)
Imprecision is often quantified in terms of the half-width of
the interval, i.e., the distance between
the point estimate and the interval’s upper (or lower) limit,
which we will refer to here as the
“margin of error”. The half-width of the above interval is 0.1
(i.e., the square root of n) in absolute
terms or 20% (0.1/0.5) in relative terms. A 0.1 absolute or
20% relative margin of error is adequate
for a “ballpark” estimate of a proportion, but not much more.
Page 41
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 491
rev. 3/29/2004, 6/27/2004, 7/22/2004
Since the above expressions involve the square root of the
sample size, progressive narrowing of the
interval width involves substantially greater increases in
sample size. For example, to obtain a 0.05
absolute or 10% relative margin of error, sample size must be
quadrupled, to 400:
(0.5 – 1/√ 400 , 0.5 + 1/√ 400 )
=
(0.5 – 1/20 , 0.5 + 1/20)
=
(0.5 – 0.05, 0.5 + 0.05)
=
(0.45, 0.55)
Similarly, a sample size of 900 yields confidence limits one-
third as wide as from a sample of 100, a
sample of 2,500 yields limits one-fourth as wide as for
n=100, etc.
These numbers apply to a point estimate of 0.5, which
produces the widest error margin in absolute
terms. A smaller or greater point estimate will have a
narrower (in absolute terms) interval, because
the square root of p(1 – p) cannot exceed 0.5 (try it! – or use
calculus). The relative margin of error,
on the other hand, is inversely related to the size of the point
estimate. Examine the following table:
Margin of error (rounded)
Point
estimate
Sample
size
Absolute * Relative **
(%)
0.1
100
0.06***
60***
0.2
100
0.080
40
0.3
100
0.090
30
0.4
100
0.096
24
0.5
100
0.100
20
0.6
100
0.096
16
0.7
100
0.090
12
0.8
100
0.080
9.8
0.9
100
0.060
6.5
0.1
400
0.030
30
0.2
400
0.040
20
0.3
400
0.045
15
0.4
400
0.048
12
0.5
400
0.050
10
0.6
400
0.048
8.0
0.7
400
0.045
6.4
0.8
400
0.040
4.9
0.9
400
0.030
3.2
* Approximate half-width of 95% confidence interval in
absolute terms
** Approximate half-width of 95% confidence interval in
absolute terms, relative to the size of
the point estimate
Page 42
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 492
rev. 3/29/2004, 6/27/2004, 7/22/2004
*** Calculation: 1.96 √[(0.01)(1– 0.01) / 100] = 1.96 (0.03)
= 0.0588 ≈ 0.06 absolute error
margin
This table illustrates that:
1. quadrupling sample size halves the margin of error.
2. absolute error margin decreases as the point estimate
moves away from 0.5
3. relative error margin is inversely – and very strongly –
related to the size of the point
estimate
For very small point estimates, as illustrated in the following
table, very large samples are required to
obtain a small relative margin of error. Even a sample size of
2,500 still produces a relative error
margin of 17% for a proportion of 0.05.
Margin of error (rounded)
Point
estimate
Sample
size
Absolute *
Relative *
(%)
0.5
100
0.10
20
0.5
400
0.05
10
0.5
900
0.033
6.6
0.5
1,600
0.025
5.0
0.5
2,500
0.020
4.0
0.05
100
0.043
85
0.05
400
0.021 **
43 **
0.05
900
0.014
28
0.05
1,600
0.011
21
0.05
2,500
0.009
17
* See previous table
** Calculation: 1.96 × √[(0.05)(0.95)/400] = 1.96 × 0.0109
= 0.0214 ≈ 0.021 absolute error margin
Relative = 0.0214 / 0.05 = 0.427 = 42.7% (approximately
43%)
Recall that this formula requires that nP > 5, which is just
met for P=0.05 and n=100.
How large a sample is large enough? If the objective is to set
an upper or lower bound on a
proportion, then a small absolute margin of error may
suffice. For example, if one is testing for
hepatitis C antibody and wants to be reassured that the
seroprevalence is below 5%, then a sample
Page 43
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 493
rev. 3/29/2004, 6/27/2004, 7/22/2004
size of 900 will produce an interval with an absolute error
margin no wider than 0.033 (for a point
estimate of 0.5 – see above table) and more likely 0.011 (for
a point estimate of 0.05) or smaller.
Since we expect the seroprevalence to be very small, then the
0.011 is much more relevant than the
0.033. If when we carry out the study we obtain a point
estimate of exactly 0.05, then the 95%
confidence interval will be (0.039,0.061) which will tell us
that the true value is at least not likely to
be greater than 6%. If the point estimate is below 0.04, then
the upper confidence limit will be
below 5% and we are reassured that the seroprevalence is no
greater than that value.
Note that the above is all based on the assumption of perfect
(unbiased) simple random sampling
and measurement. Anything other than an unbiased simple
random sample and any error in
measurement will invalidate the above at least to some
extent.
Meditations on hypothesis testing and statistical significance
The statistical theory of hypothesis testing and assessment of
statistical “significance” proceeds from
an analysis of decision-making with respect to two
competing hypothesis: a “null” hypothesis and
an alternative hypothesis. Two types of errors are possible:
Type I: Erroneously reject the “null hypothesis” (H0), in
favor of the alternate hypothesis (HA), i.e.,
erroneously reject chance as a sufficient explanation for the
observed results.
Type II: Erroneously fail to reject H0, i.e., erroneously
accept chance as an explanation. [A parallel
dichotomy will be seen later in the course when we discuss
sensitivity and specificity.]
Traditionally, the Type I error probability has received more
attention and is referred to as the
“significance level” of the test. The Type I error presumably
owes its prominence to the scientific
community’s desire to avoid false alarms, i.e., to avoid
reacting to results that may readily have been
chance fluctuations. Also the Type I error probability is
easier to estimate, since the Type II error
probability depends on stating the size of true difference that
one seeks to detect.
During recent decades, the calculation and presentation of
“p-values” (which give information about
the Type I error probability) have become de rigueur in the
empirical scientific literature. Indeed, a
significant (!) number of people refuse to pay any attention to
results that have p-values greater than
.05 (5% probability of a Type I error).
Such a stance is a considerable labor-giving device, but is
perhaps a bit brutal. After all, a result with
a p-value of .10 would result from a chance process in only
one in ten trials. Should such a finding
be dismissed? Moreover, since the p-value reflects the
number of subjects as well as the size of the
observed difference, a small study will have very small p-
values only for large (and perhaps
unrealistic?) observed differences. If the size of the observed
difference is unreasonably large, then
we may be suspicious of the finding despite a small p-value.
If the observed difference is plausible,
but because the study is small the p-value is “not
significant”, we may nevertheless want to pay some
attention.
Page 44
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 494
rev. 3/29/2004, 6/27/2004, 7/22/2004
Another reason for a reflective, rather than a reflexive,
approach to p-values (and statistical inference
generally) is that the probability estimates themselves are
accurate only with respect to the models
that underlie them. Not only may the mathematical models
not adequately capture the real situation,
but the context in which they are used clouds the issue. One
critical assumption is that of random
sampling or randomization (as in a randomized controlled
trial). Although this assumption is the
basis for virtually all of the statistical theory of hypothesis
testing and confidence intervals, it is rarely
met in observational studies and the limitations that it
imposes on the interpretation of statistical
tests are often underappreciated (Greenland S.
Randomization, statistics, and causal inference
Epidemiology 1990;1:421-249).
Even in randomized trials problems of interpretation exist.
For example, the p-value for a single
result in a single study may be 5 percent. But that means that
20 independent studies of two
identical phenomena would observe, on the average, one
difference that was “significant” at the five
percent level. A prolific investigator who conducts 200
studies in his/her professional life can
expect to have ten that are “significant” by chance alone.
Moreover, a study will often examine
multiple outcomes, including multiple ways of defining the
variables involved.
Such “multiple comparisons” increase the likelihood of
chance differences being called “significant”.
But the statistical procedures for dealing with this
“significance inflation” tend, like measures to
suppress price inflation or grade inflation, to produce
recession or even depression [of study
findings]. Should an investigator be required to take an oath
that he/she had (1) fully specified an a
priori hypothesis, including the procedures for defining and
manipulating all variables, decisions
about all relationships to examine, what factors to control,
etc; (2) proceeded directly to the pre-
specified statistical test without looking at any other data;
and (3) will not perform any further
statistical tests with the same data? (See Modern
Epidemiology for more on these points.)
What about so called “fishing expeditions” in which an
investigator (or her computer) pores over a
dataset to find “significant” relationships. Is such a procedure
better characterized as “seek and ye
shall find” or as “search and destroy”? Some analysts
recommend adjusting the significance level to
take account of such “multiple comparisons”, but an
energetic investigator can easily perform
enough tests so that the adjusted significance level is
impossible to satisfy. Other writers (e.g.,
Rothman, Poole) assert that no adjustment is required – that
once the data are in, the number of
tests is irrelevant. Others (e.g., Greenland) have proposed
more sophisticated approaches to
adjustment. Perhaps the best course at this time is twofold:
(1) If you are conducting a study, for example, a randomized
trial, in which you have a good chance
of satisfying the assumptions for a statistical hypothesis test
and are hoping to test a specific
hypothesis, especially one that may lead to some decision,
then it is probably better to adhere to the
Neyman-Pearson hypothesis testing format as much as
possible. This approach ensures maximum
impact for your results;
(2) If you are conducting an inquiry with few of the above
characteristics, or have already
completed the a priori hypothesis test, analyze all that you
like but be candid in describing how you
proceeded. Then readers can interpret as they judge most
appropriate.
Page 45
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 495
rev. 3/29/2004, 6/27/2004, 7/22/2004
Apparent (calculated) power is rarely achieved because it
often assumes no errors in classification of
subjects. A study with advertised power of 90% could well
have much less probability of detecting a
given true difference because of dilution by information bias.
Similarly we can in principle improve
the effective power of a study if we can increase the
precision with which important variables are
measured.
Louis Guttman has written that estimation and
approximation, never forgetting replication, may be
more fruitful than significance testing in developing science.
[Louis Guttman. What is not what in
statistics. The Statistician 25(2):81-107.]
Independent replication is the cornerstone of scientific
knowledge.
Bayesian approach to p-value interpretation
The application of the concepts of sensitivity, specificity, and
predictive value to interpreting
statistical hypothesis tests suggests an analogy between
statistical tests and diagnostic tests (see
Browner and Newman, 1987; Diamond and Forrester, 1983;
and Feinstein, Clinical Biostatistics). Just
as the interpretation of a diagnostic test depends upon disease
prevalence (the “a priori likelihood
that the patient has the disease”), the interpretation of
statistical tests can be regarded as dependent
upon “truth prevalence”, i.e., on the reasonableness of the
hypothesis.
As noted earlier, we would like statistical inference to
provide an estimate of the probability that a
hypothesis of interest (H) is true given the observed results.
The p-value provides instead the
probability of observing an extreme result under a null
hypothesis (typically the inverse of the
hypothesis of interest). The Bayesian approach to
interpreting p-values tries to provide an answer
that comes closer to the original objective. In the Bayesian
approach, we begin with a prior
probability for the truth of the hypothesis and then adjust that
probability based on the results of a
study, to obtain a posterior probability. The effect that the
study results have on our assessment of
the credibility of the hypothesis depends on our original
assessment of its credibility.
Let T mean that a statistical test is “significant”. According
to Bayes Theorem, if Pr(H) is the “a
priori” probability of H, i.e., the probability that H is true
based only on previous information, then
the a posteriori probability of H (the probability that H is true
based on previous information and the
current test result) is:
Pr(H) Pr(T|H)
Pr(H|T) = ––––––––––––––––––––––––––––
Pr(H) Pr(T|H) + Pr(h) Pr(T|h)
[where Pr(T|h) means the probability of a positive test given
that the hypothesis is not true] which
can be written:
Page 46
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 496
rev. 3/29/2004, 6/27/2004, 7/22/2004
Pr(H) Pr(T|H)
Pr(H|T) = –––––––––––––––––––––––––––––––
Pr(H) Pr(T|H) + [1 – Pr(H)] Pr(T|h)
Since Pr(T|H) is the statistical power (the probability of a
positive test given that the hypothesis is
true) and Pr(T|h) is the p-value (the probability of a positive
test given that the hypothesis is not
true), the posterior probability can be written:
Pr(H) (power)
Pr(H|T) = –––––––––––––––––––––––––––––––
Pr(H) (power) + [1 – Pr(H)] (p-value)
Pr(H|T) is therefore a function of the “a priori” probability of
the hypothesis, the statistical power,
and the p-value. Therefore the p-value has more impact on
Pr(H|T) when Pr(H) is small (i.e., when
a hypothesis is not supported by prior research or laboratory
data) (see Diamond and Forrester).
To get an idea of how these formulas work with typical
values for the various elements, take a look
at the following table:
Page 47
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 497
rev. 3/29/2004, 6/27/2004, 7/22/2004
Evaluation of posterior probability based on
prior probability, statistical power, and p-value
Prior
probability
(Before
the study)
Statistical
power
of the
study
P-value
(Findings
of the
study)
Posterior
probability
(After the
study)
Pr(H)
Pr(T|H)
Pr(T|h)
Pr(H|T)
Credible
0.60
0.8
0.100
0.92
hypotheses
0.60
0.8
0.050
0.96
High power
0.60
0.8
0.001
1.00
0.60
0.5
0.100
0.88
0.60
0.5
0.050
0.94
Low power
0.60
0.5
0.001
1.00
Long shot
0.05
0.8
0.100
0.30
hypotheses
0.05
0.8
0.050
0.46
High power
0.05
0.8
0.001
0.98
0.05
0.5
0.100
0.21
0.05
0.5
0.050
0.34
Low power
0.05
0.5
0.001
0.96
In this table, for example, a very strong p-value (e.g., 0.001)
gives high credibility (posterior
probability) even for a long shot hypothesis examined in a
study of low statistical power. A p-value
that is “just significant”, however, does not make a
hypothesis highly credible unless it was judged
more likely than not before the study. Even a
“nonsignificant” p-value (e.g., 0.10) provides some
increase in credibility of the hypothesis, so in the Bayesian
framework a p-value of 0.10 would not
be regarded as a “negative” result casting doubt on the
existence of an association. Meta-analysis, in
which results are combined across studies to obtain a
quantitative assessment of an association from
the full body of evidence, also takes into account evidence
for the association from studies that
observed an association but had a p-value greater than 0.05.
Formal use of Bayesian methods in
everyday work, however, is somewhat constrained by the
absence of an obvious method for
obtaining a prior probability.
We can also use these concepts to show why the frequent
temptation to interpret a small p-value as
the probability that the result was due to chance is mistaken.
When confronted with a p-value of,
e.g., 0.05, many statistical novices are tempted to interpret
the result as meaning that “the probability
that the finding was due to chance is only 5%”, which is
equivalent to a posterior probability that the
hypothesis is true of 1 – 0.05 = 0.95 = 95%. From the above
table, we see that the only situations in
which the posterior probability is close to one minus the p-
value occur when the hypothesis is more
likely than not to be true (a priori probability of 0.60). For
long shot hypotheses, a p-value of 0.05
corresponds to a posterior probability of much less than 95%.
Page 48
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 498
rev. 3/29/2004, 6/27/2004, 7/22/2004
More meditations on interpreting statistical significance tests
Some of the concepts in the interpretation of statistical tests
of significance can perhaps be
illustrated through an example based on one glorious origin
of probability theory – games of chance.
Suppose that our friend tells you that he has an intuition
about roulette wheels. By watching the
operator spin the wheel, your friend can, he claims, predict
where the ball will land within a very
small margin. If, for simplicity, the wheel has numbers 1-100
on it, your friend says he can predict
the numbers where the ball will land. He wants you to put up
some money to send him to Monte
Carlo to make our fortunes.
Naturally you’re excited by the prospect of instant riches but
also a bit skeptical. To verify your
friend’s claim, you undertake a statistical test. You give your
friend $5 to prove his prowess at the
local gambling casino, and you wait to see how he does.
The null hypothesis for your statistical test is that your friend
has no special ability, so that his
chances of predicting the resting place of the ball on any one
try are simply 1 out of 100 (.01). The
1-sided alternate hypothesis is that your friend does have this
ability and can predict the correct
number more often than 1 out of 100 times. [The 2-sided
alternate hypothesis is that your friend
will predict the resting place of the ball either more than
would be expected by chance or less than
would be expected.]
Your friend returns with $400. Knowing that the probability
of his being correct on a given try by
chance alone was only 1%, your are impressed. His
performance was “significant at the .01 level”!
Do you underwrite his trip to Monte Carlo? How do you
interpret his correct prediction?
Is it correct to say that there is only a 1% chance that his
accurate prediction was due to “luck”?
Not quite. According to the frequentist interpretation, the
prediction was made and the roulette
wheel has already been spun. The accuracy was due either to
“chance” (“luck”) or your friend’s
ability, but only one of these was actually responsible that
time. So the probability that his correct
prediction was due to chance is either zero (i.e., your friend
can predict) or one (your friend cannot
predict). The only trouble is, you don’t know which!
You can say (before the wheel was spun and assuming it was
a balanced wheel) that if your friend
had no special ability there was only a one percent
probability of his making a correct prediction and
that therefore his winning is evidence against the null
hypothesis (of no ability) and in favor of the
alternate hypothesis (ability to predict). If you have to decide
that day, you might figure that it
would be worth underwriting his trip to Monte Carlo, but you
would be aware that his correct
prediction could have been due to chance because there was
a one percent probability that in the
absence of any clairvoyance his prediction would have been
correct (not quite the same as a one
percent probability that his correct prediction was due to
chance). So you give your friend $2,000.
He thanks you profusely, and in parting, tells you that it
actually took him 30 tries to make a correct
prediction – he borrowed the money for the other 29 tries.
Page 49
_________________________________________________
____________________________________________
www.epidemiolog.net © Victor J. Schoenbach
14. Data analysis and interpretation – 499
rev. 3/29/2004, 6/27/2004, 7/22/2004
That information gives you pause. Certainly you would not
have been so impressed if he had told
you he could make a correct prediction in 30 tries. If the
probability of a correct prediction (i.e., a
correct guess) in the absence of any special ability is 0.01,
then the probability of one or more
correct guesses in 30 tries is 0.26 (1.0 minus the quantity
0.99 raised to the 30th power). Twenty-six
percent is still less than 50 percent, i.e., the probability of
winning a coin flip, but not so
impressively. The evidence against the null hypothesis is
now not nearly so strong. This change in
your interpretation illustrates the issue that arises in
connection with multiple significance tests and
small studies bias.
It is possible, using statistical theory, to adjust significance
levels and p-values to take into account
the fact that multiple independent significance tests have
been done. But there are various practical
problems in applying such procedures, one of which is the
lack of independence among multiple
tests in a particular set of data. For example, if your friend
explained that he so rarely makes an
incorrect prediction that when he did he became so upset that
it took him a whole hour (and 29
more predictions) to regain his predictive ability, then even if
you remained skeptical you would be
hard-put to calculate an adjusted p-value for your test if you
thought he was telling the truth.
Similarly, in a given dataset, does the fact that an investigator
tested the same difference in various
ways (e.g., obesity as indexed by weight/height2 [Quetelet’s
index], weight/height3 [ponderal index],
percent above ideal weight, skinfold thickness, and body
density) weaken the findings for each test?
If she also looked at blood pressure differences, would that
weaken the credibility of statistical
significance of differences in obesity?
“You pays your money, and you takes your choice.”
DATA INTERPRETATION
The application of statistics in research is well documented.
Before choosing a statistical method for your own research
project, knowledge regarding scales of measurement is a
prerequisite. Scales of measurement per se have to do with
the allocation of numerical values to characteristics
according to certain rules. Measurement can thus either be
quantitative or qualitative. The quantitative level of
measurement includes among other things, aspects such as
interpretation and paragraph analysis, whilst the quantitative
level of measurement focuses on measures such as nominal,
ordinal, internal and ratio levels of measurement. The latter
are basic scales of measurement and will be briefly outlined.

1 Measuring scales

* Nominal measurement
Nominal measurement includes the awarding of a numeral
value to a specific characteristic. Tr~is type of measurement
is the most basic form of measurement, because it measures
the lowest level that can be measured and is therefore
considered a scale of measurement with limitations. The
following serves as an example of nominal measurement: A
researcher wants to determine the profile of the academic
background of his students. For this he/she might need
information regarding the specific level (HG, SD, LG) his
students passed during their matriculation examination. The
different matriculation subjects would then be listed i.e.
Mathematics, English, Geography etc. To each subject
passed on the higher grade a numerical value 1 will be
allocated, and the numerical value 2 for a pass in the standard
grade and so forth. The numerical value 1 does not mean half
of 2, it merely indicates a difference.

This scale of measurement is used mainly for the compiling

of frequency tables (Smit 1983: 208).

* Ordinal measurement

Ordinal measurement is applicable in cases where a

criterion/characteristic is awarded to numeral value in terms
of a specific order (thus the name of the scale).

The ordinal scale implies that the entity being measured is

quantified in terms of either higher or lower, greater (more)
or lesser (less) without specifying the size of the intervals
(Leedy 1993: 38). The numerical 1 can be the highest, whilst
3 could be the lowest. The following serves as an example of
measurement on ordinal level: A school wants to select the
best student of the year. After the evaluators (in this case the
teachers) have nominated the best students, the finalists
should be placed in rank order according to the set criteria in
order to determine the best student. In this case the numeral 1
is an indication of the best student, and therefore the winner.
The ordinal scale of measurement can be applied to
determine the median, percentage, rank order, correlations
and percentile (Smith 1983: 209).

* Interval measurement

The interval scale of measurement is characterized by two

features, namely:

- equal units of measurement (equal intervals); and

- a zero point which has been established arbitrarily (Leedy

1993: 38).

The latter indicates that there is not an absolute zero point.

There is therefore a specific: relationship between the
distance (interval) of the numerical value and the different
sizes of a characteristic. Because of the before mentioned
characteristics, this measure scale is considered to be a more
advanced type of measuring scale. An increase or decrease of
the one characteristic goes hand in hand with an increase or
decrease of the other. The interval level of measurement
enables the researcher to compromise between aspects and to
indicate clearly how much more the one has of a
characteristic than the other. As an example of this type of
measurement the following is given: an individual
intelligence score is 110 whilst another is 100. The difference
between these scores is exactly 10. It should also be
remembered that intelligence measurement has no zero point.
A second example of interval measurement is the Lickert
attitude scale.

The interval scale of measurement is therefore suitable to

calculate arithmetic mean averages, do standard deviations
and determine correlation studies, provided that the
researcher takes care that the preconditions set for each scale
of measurement, are abided by.

* Ration measurement

This is considered the highest order of measurement that

exist, because of the fixed proportions (ratio) between the
number (numerical) and the amount of the characteristic; that
it represents. What should be mentioned is that, when ration
levels are measured, a fixed (absolute) zero point exists.
Ration level of measurement thus enables researchers to
determine whether aspects possess something of a
characteristic or not. The following can serve as example of
ration level of measurement: The average weight of a
gymnast is 55 kilograms. On the other hand, the average
weight of a rugby player is 75 kilograms. Kilograms are
expressed in constant units, and a zero point does exist,
because "no weight" can be determined. Because scores on
this type of measuring scale possess absolute values, any
type of arithmetical calculations are allowed.

2 Characteristics of measuring scales

With any type of measurement, two considerations are

important - validity on the one hand and reliability on the
other hand.

* Reliability

Reliability is the term used to deal with accuracy. A scale

measurement is considered reliable if it measures that which
it is supposed to measure. Further refining of the term
reliable is that, when a test is repeated by the same researcher
i.e. with a different group representing the original group, the
same results should be obtained .
 * Validity

Validity on the other hand is concerned with the soundness

and the effectiveness of the measuring instrument (Leedy 1
993: 40).

From the literature consulted four types of validity stands

out, namely:

- content validity,

- prognostic validity,

- simultaneous validity; and

- construct validity.

Content validity deals with the accuracy with which an

instrument measures the factors or content of the course or
situations, under study.

Prognostic validity on the other hand relies basically upon

the possibility to make judgements by virtue of results
obtained by the instrument. The judgement is future
orientated. Consider the following as an example: it can be
predicted from the matriculation results of a prospective
student that he would be a successful medical student.

Simultaneous validity is tested by comparing one measuring

instrument to another one that measures the same
characteristic and which is available immediately. The
second, as a criteria, checks the accuracy of the first measure
and sets a standard against which to measure the results. The
data of the measuring instrument should correlate with
equivalent data of the criterion.
Construct validity is interested in the degree to which the
construct is actually measured.

3 Classification of statistical methods

Before a researcher can use a statistical method for his

research, he should be familiarized with the various
statistical methods as well as the prerequisites for the
implementation thereof. Because of the circumference of
statistical methods, an in-depth discussion cannot take place
for the purpose of this element. It will suffice to highlight the
basic statistical methods.

Statistical methods in the broadest sense are classified into

two main groups namely descriptive and inferential statistics.

* Descriptive statistics
Smit (1983: 212) sees descriptive statistics as the formulation
of rules and procedures according to which data can be
placed in useful and significant order. Landman (1988: 94)
states that descriptive statistics deals with the central
tendency, variability (variation) and relationships
(correlations) in data that are readily at hand. The basic
principle for using descriptive statistics is the requirement for
absolute representation of data.

The most important and general methods used are:

- Ratios. This indicates the relative frequency of the

various variables to one another, for example 1.

- Percentages. Percentages (%) are calculated by

multiplying a ratio with 100. In other words it is a ratio that
represents a standard unit of 100.
- Frequency tables. It is a means to tabulate the rate of
recurrence of a specific measurement, for example a specific
achievement in a test. Data arranged in such a manner is
known as distribution. If the distribution tendency is large,
larger class intervals are used in order to acquire a more
systematic and orderly system.

In order to understand and interpret a frequency table, you

are referred to Huysamen (1976: 24).

* The histogram. The histogram is a graphic

representation of frequency distribution and is being used to
represent simple frequency distribution. Characteristic is a
vertical line (the y axis/ordinate) at the left sideline of the
figure and the horizontal line (x axis) at the bottom. The two
lines meet at a 90 grade angle.

Because frequencies should be divided into class

intervals, the benefit of graphic presentation is that data can
be observed immediately.

* Frequency polygon. The frequency polygon does not

differ basically from the histogram, but is only used for
continual data. Instead of drafting bars for the complete
histogram, a dot indicating the highest score is placed in the
middle of the class interval. When the dots are linked up, the
frequency polygon is formed. Usually an additional class is
added to the end of the line in order to form an anchor.

* Cumulative frequency curve. The frequency on the

frequency table is added, starting from the bottom of the
class interval, and adding class by class. The cumulative
frequency in a specific class interval can then clearly indicate
how many persons/ measurements perform below or above
the class intervals. In other words, from cumulative
frequency tables a curve can be drawn, to reflect data in a
graphic manner (Smit 1983: 213).

wpe3.jpg (18580 bytes)

* Percentile curve. The cumulative frequency can also

be converted into percentages or proportions of distribution.
From such a table, one could read certain percentages or
promotions of persons or cases, with regards to a certain
point on the scale. The scale value in which 10% of the score
in a distribution falls, is regarded as the P10 (10 percentile).
Those in which 25% of the score falls is the first quarter of
P25 etc. (Smit 1983: 213).

wpe4.jpg (15279 bytes)

* Line graphic. During the previous graphic

presentations the historical line (X axis) indicated the scale
of measurement, whilst the vertical line (Y axis) indicated
the frequency. In the case of a line graphic, both axes (X and
Y) are used to indicate the scale of measurement with the
aim of indicating a comparison between two comparable
variables (Smit 1983: 214).

wpe5.jpg (13539 bytes)

4 Central tendency

Central tendency is defined as the central point around which

data revolve. The following techniques can be employed:

* The mode

The mode is defined as the score (value or category) of the

variable which is observed most frequently. For example:

375864595
From the above mentioned, the mode equals 5 because 5
appears to be the most frequent score amongst all the
numbers (occurred 3X).

* Median

The median indicates the middle value of a series of

sequentially ordered scores. Because the median divides
frequencies into two equal parts, it can also be described as
being the fiftieth percentile.

10 13 14 15 18 19 22 25 25

The median in the above-mentioned is the fifth score, that is

18. There are 4 counts on both sides of the numerical value
18.
In cases where you have, for instance:

10 13 14 15 18 19 22 25 26 29

there are 2 numerical values indicating the median. By

dividing the result by 2, the median can be determined. The
fifth score with a numerical value of 18 and the sixth score
with the numerical value of 19 are in the middle of the
sequentially ordered scores. The median for the above
mentioned scores is therefore 18 + 19 ) 2 = 18,5. Because
18,5 does not occur in the sequentially ordered scores,
Huysamen (1983: 50) states that one should in cases of these
rather refer to the 18.5 percentile.

* Arithmetic mean

The arithmetic mean refers to a measure of central tendencies

found by adding all scores and dividing them by the number
of scores. The following is an example:

5 2 6 1 6 = (Sum total of scores wpe7.jpg (788

bytes))

Thus 5 + 2 + 6 + 1 + 6 = 20, because there are 5 scores, N =

5, and the sum total of the scores (20) is divided by 5.

* Standard deviation

The standard deviation is a measure of the spread of

dispersion of a distribution of scores. The deviation of each
score from the mean is squared; the squared deviations are
then summed, the result divided by N-1, and the square root
taken (Landman 1988: 94).
 * Inference statistics

Apart from descriptive statistics that deal with central

tendencies, statistical methods enabling researchers to go
from the known to the unknown data also exist. This is to say
to make deductions or statements regarding the broad
population as the samples from which the 'known' data are
drawn. These methods, according to literature are called
inferential or inductive statistics (Landman 1988: 95). These
methods includes estimation, predictions, hypothesis testing
and so forth.

In conclusion the role of statistical methods in research is to

enable the researcher to accurately utilize the gathered
information and to be more specific in describing his
findings. For more details on statistical calculations you are
referred to Huysamen (1976).
5 Self-test

a) The following marks were allocated to students during a

test they wrote:

33 44 69 66 72 46 69

44 61 80 73 42 73 88

62 81 75 50 71 56 86

60 86 54 80 87 63 49

Compile the following on the scores presented to you:

- frequency distribution

- histogram
 - frequency polygon.

b) Calculate the arithmetic mean, mode and median of the

following:

8 7 3 21 16 34 22 18 19.

c) Determine the mode and median of the following:

8 11 12 3 31 12 8 9 12 10 5.

6 SOURCE LIST

Huysamen, GK 1976 Beskrywende Statistiek vir Sosiale

Wetenskappe. Pretoria: Academica.

Landman, WA 1988 Navorsingsmetodologiese

Grondbegrippe. Pretoria: Serva.

Leedy, PD 1985 Practical Research: Planning and

Design. Third Edition. New Tork: McMillan Publishing Co.

Smit, GJ 1983 Navorsingsmetodes in die

geesteswetenskappe. Pretoria: HAUM.

7 ANSWERS TO SELF TEST

b wpe6.jpg (788 bytes) = 15

Modus = 21

Median = 16

c Modus = 12
 Median = 12
Use of Computers in Accounting

Introduction

Computers play an important part in the recording of finical

information.

There care many accounting packages available, and so

many businesses
are able to use computerised accounting system.

One of the most important factors of computerised

accounting system is
that it provides the same functions as a manual accounting
system.

Main body
Most business use computer systems instead rather then
manual systems
to record finical information, because it is a lot faster file can
be
shared more easily and changes can be made easily.

Tasks that can be performed using computers are

spreadsheets, which
are used for a variety of functions:

· Producing invoices - working out the costs of products sold,

calculating and adding VAT and producing sales total.

· Working out budgets for future expenditure.

· Working out sales figures for different products or areas.

Most accounting packages will provide the basis for
daybooks, ledger
accounts, (double entry), tri.

"THE ROLE OF COMPUTERS IN MARKETING"

Marketing is the process by which goods are sold and
purchased. The aim of marketing is to acquire, retain, and
satisfy customers. Modern marketing has evolved into a
complex and diverse field. This field includes a wide variety
of special functions such as advertising, mail-order business,
public relations, retailing and merchandising, sales, market
research, and pricing of goods.

Businesses, and particularly the marketing aspect of

businesses, rely a great deal on the use of computers.
Computers play a significant role in inventory control,
processing and handling orders, communication between
satelite companies in an organization, design and production
of goods, manufacturing, product and market analysis,
advertising, producing the company newsletter, and in some
cases, complete control of company operations.

In today's extremely competitive business environment

businesses are searching for ways to improve profitability
and to maintain their position in the marketplace. As
competition becomes more intense the formula for success
becomes more difficult. Two particular things have greatly
aided companies in their quests to accomplish these goals.

They are the innovative software products of CAD/CAM

and, last but not least, the World Wide Web.

An important program has aided companies all over the

world. Computer-aided design and computer-aided
manufacturing (CAD/CAM) is the integration of two
technologies. It has often been called the new industrial
revolution. In CAD, engineers and designers use specialized
computer software to create models that represent
characteristics of objects.

These models are analyzed by computer and redesigned as

necessary. This allows companies needed flexibility in
studying different and daring designs without the high costs
of building and testing actual models, saving millions of
dollars. In CAM, designers and engineers use computers for
planning manufacturing processes, testing finished parts,
controlling manufacturing operations, and managing entire
plants. CAM is linked to CAD through a database that is
shared by design and manufacturing engineers.

The major applications of CAD/CAM are mechanical design

and electronic design. Computer-aided mechanical design is
usually done with automated drafting programs that use
interactive computer graphics. Information is entered into the
computer to create basic elements such as circles, lines, and
points. Elements can be rotated, mirrored, moved, and
scaled, and users can zoom in on details. Computerized
drafting is quicker and more accurate than manual drafting. It
makes modifications much easier.

Desktop manufacturing enables a designer to construct a

model directly from data which is stored in computer
memory. These software programs help designers to consider
both function and manufacturing consequences at early
stages, when designs are easily modified.

More and more manufacturing businesses are integrating

CAD/CAM with other aspects of production, including
inventory tracking, scheduling, and marketing. This idea,
known as computer-integrated manufacturing (CIM), speeds
processing of orders, adds to effective materials
management, and creates considerable cost savings.

In addition to designing and manufacturing a product, a

company must be effectively able to advertise, market, and
sell its product. Much of what passes for business is nothing
more than making connections with other people. What if
you could passout your business card to thousands, maybe
millions of potential clients and partners? You can, twenty
four hours a day, inexpensively and simply on the World
Wide Web. Firms communicate with their customers through
various types of media. This media usually follows passive
one-to-many communication where a firm reaches many
current and potential customers through marketing efforts
that allow limited forms of feedback on the part of the
customer. For several years a revolution has been developing
that is dramatically changing the traditional form of
advertising and communication media. This revolution is the
Internet, a massive global network of interconnected
computer networks which has the potential to drastically
change the way firms do business with their customers.

The World Wide Web is a hypertext based information

service. It provides access to multimedia, complex
documents, and databases. The Web is one of the most
effective vehicles to provide information because of its visual
impact and advanced features. It can be used as a complete
presentation media for a company's corporate information or
information on all of its products and services.

The recent growth of the world wide web (WWW) has

opened up new markets and shattered boundaries to selling to
a worldwide audience. For marketers the world wide web can
be used to creat a client base, for product and market
analysis, rapid information access, wide scale information
dissemination, rapid communication, cost-effective
document transfers, expert advise and help, recruiting new
employees, peer communi- cations, and new business
opportunities. The usefullness of the Internet or WWW
depends directly on the products or services of each business.
There are different benefits depending upon the type of
business and whether you are a supplier, retailer, or
distributor. Lets examine these in more detail.

Finding new clients and new client bases is not always an

easy task. This process involves a careful market analysis,
product marketing and consumer base testing. The Internet is
a ready base of several million people from all walks of life.
One can easily find new customers and clients from this
massive group, provided that your presence on the internet is
known. If you could keep your customer informed of every
reason why they should do business with you, your business
would definitely increase. Making business information
available is one of the most inportant ways to serve your
customers. Before people decide to become customers, they
want to know about your company, what you do and what
you can do for them. This can be accomplished easily and
inexpensively on the World Wide Web.

Many users also do product analyses and comparisons and

report their findings via the World Wide Web. Quite
frequently one can find others who may be familiar with a
product that you are currently testing. A company can get
first hand reports on the functionality of such products before
spending a great deal of money. Also, the large base of
Internet users is a principle area for the distribution of
surveys for an analysis of the market for a new product of
service ideas. These surveys can reach millions of people and
potential clients with very little effort on the part of the
surveyors. Once a product is already marketed, you can
examine the level of satisfaction that users have received
from the product. Getting customer feedback can lead to new
and improved products. Feedback will let you know what
customers think of your product faster, easier and much less
expensively than any other market you may reach. For the
cost of a page or two of Web programming, you can have a
crystal ball into where to position your product or service in
the marketplace.

Accessing information over the Internet is much faster on

most occasions than transmissions and transfers via fax or
postal courier services. You can access information and data
from countries around the world and make interactive
connections to remote computer systems just about anywhere
in the world. Electronic mail has also proved to be an
effective solution to the problem of telephone tag. Contacting
others through email has provided a unique method of
communication which has the speed of telephone
conversations, yet still provides the advantages of postal
mail. Email can be sent from just about anywhere that there
is an Internet service or access so that businessmen or
travelers can keep in touch with up to the minute details of
the office.

Another benefit of the World Wide Web is wide scale

information circulation. You can place documents on the
Internet and instantly make them accessible to millions of
users around the world. Hypertext documents provide an
effective technique by which to present information to
subscribers, clients or the general public. Creating World
Wide Web documents and registering your site with larger
Web sites improves the availability of the documents to a
client base larger, and cheaper, than the circulation of many
major newspapers and/or television medias. You may not be
able to use the mail, phone system and regulation systems in
all of your potential international markets. With the World
Wide Web, however, you can open up a dialogue with
international markets as easily as with the company accross
the street.
The Web is also more cost-effective than conventional
advertising. Transferring on- line documents via the Internet
takes a minimal amount of time, saving a great deal of
money over postal or courier services which can also suffer
late deliveries, losses or damage. If a document transfer fails
on the Internet, you can always try again since the cost of the
transfer is exactly the same. Current or potential clients are
not lost due to late or absent documents.

Beyond product and market analysis, there are a great

number of experts on the Internet who make their presence
widely known and easily accessable. Quite often you can get
free advice and help with problems you might have from the
same people you may otherwise pay highly for their
consulting services to large organizations, magazines, and
other periodicals. Researchers and business executives alike
have attested to the fact that much of their communications
over the Internet are with others in their line of research or
field of work. Communicating with peers allows the sharing
of ideas, problems and solutions among themselves. Often
people find that others in their field have already created
solutions for problems similar to their own. They are able to
obtain advice on their own situations and create a solution
based upon this shared knowledge.

Many businessmen and conpanies are continuously on the

look-out for new and innovative ideas as viable business
ventures. Internet users are continuously coming up with
such new ideas because of the available research the Internet
offers and also because of the cooperative atmosphere that
surrounds the internet. In addition, the Internet has many job
lists and resumes online for prospective employers. New
resumes are constantly posted to the Web to inform
companies of the availability of new skills.
As competition intensifies in the business world, consumers
are faced with more and more products and services to
choose from. The future of business is being decided right
now in the minds and wallets of customers. The successful
business and marketing approach utilizes everything possible
to insure that the choice the customer makes is to choose
their product or service. Computer technology is by far the
most important and impressive means by which to insure a
company's success. Computers play a significant role in
every aspect of a company's survival, from product design
and manufacturing, creating client databases, inventory
control , market analysis, advertising and sales, and even
total company operations

Role Of Computers In Marketing

THE ROLE OF COMPUTERS IN MARKETING

Marketing is the process by which goods are sold and

purchased. The aim of marketing is to acquire, retain, and
satisfy customers. Modern marketing has evolved into a
complex and diverse field. This field includes a wide variety
of special functions such as advertising, mail-order business,
public relations, retailing and merchandising, sales, market
research, and pricing of goods.

Businesses, and particularly the marketing aspect of

businesses, rely a great deal on the use of computers.
Computers play a significant role in inventory control,
processing and handling orders, communication between
satellite companies in an organization, design and production
of goods, manufacturing, product and market analysis,
advertising, producing the company newsletter, and in some
cases, complete control of company operations.

In today's extremely competitive business environment

businesses are searching for ways to improve profitability
and to maintain their position in the marketplace. As
competition becomes more intense the formula for success
becomes more difficult. Two particular things have greatly
aided companies in their quests to accomplish these goals.

They are the innovative software products of CAD/CAM

and, last but not least, the World Wide Web.

An important program has aided companies all over the

world. Computer-aided design and computer-aided
manufacturing (CAD/CAM) is the integration of two
technologies. It has often been called the new industrial
revolution. In CAD, engineers and designers use specialized
computer software to create models that represent
characteristics of objects.

These models are analyzed by computer and redesigned as

necessary. This allows companies needed flexibility in
studying different and daring designs without the high costs
of building and testing actual models, saving millions of
dollars. In CAM, designers and engineers use computers for
planning manufacturing processes, testing...

Narrate the role of computers in accounting. (June 03)

Computers are probably the only tool available that can help
us in storage and organization of data and information.
Today computer industry is the biggest industry worldwide
and has a great impact on the society. The computers became
popular because of the following features:

* Speed
* Reliability
* Diligence
* Versatility
* Large memory.
Role of computers in accounting & Business organizations

Help in Operations: Computers help in various accounting

operations like invoicing, calculation of wages, maintaining
ledger, etc.

Help in Controlling: With the help of computers, a business

concern can have better control over its operation. They help
in budgetary control, sales analysis, credit control, etc.

A computer can process almost any type of information

required by a business. Some areas where computers are
most popular are listed below:

* Inventory control
* System analysis
* Inventory control
 * Sales accounting
* Market research
* Purchase accounting
* Planning & control
* Quality control
* Management accounting

Q. Examine the utility of available software for performing

accounting and finance functions. Put forth your suggestions
for an ideal computerized system of accounting & finance.
(June 03)

The modern age is called the "Computer Age" or

"Information Age" because computers are becoming very
popular. People use computers in a wide variety of ways. In
business, computers track inventories with bar codes and
scanners, check the credit status of customers, and transfer
funds electronically. In homes, tiny computers embedded in
the electronic circuitry of most appliances control the indoor
temperature, operate home security systems, tell the time,
and turn videocassette recorders on and off. Computers in
automobiles regulate the flow of fuel, thereby increasing gas
mileage. Computers also entertain, creating digitized sound
on stereo systems or computer-animated features from a
digitally encoded laser disc. Computer programs, or
applications, exist to aid every level of education, from
programs that teach simple addition or sentence construction
to programs that teach advanced calculus. Educators use
computers to track grades and prepare notes; with computer-
controlled projection units, they can add graphics, sound, and
animation to their lectures. Computers are used extensively
in scientific research to solve mathematical problems, display
complicated data, or model systems that are too costly or
impractical to build, such as testing the air flow around the
next generation of space shuttles. The military employs
computers in sophisticated communications to encode and
unscramble messages, and to keep track of personnel and
supplies.

The computers have become popular because of:

* Speed
* Reliability
* Diligence
* Versatility
* Large memory.

There are many softwares available in the market for the

purpose of maintaining accounts such as TALLY, BUSSY,
Account Manager, etc. The most widely accepted is the
TALLY. TALLY provides the following functions for
accounting procedure:
 * Maintaining ledger & journals
* Preparing balance sheet, profit & loss a/c.
* Ratio Analysis
* Maintaining funds & cash flow statement
* Preparing reports

Ideal computerized system of accounting & finance must be

able to perform following functions:

* Generating reports
* Fast access of accounts
* Comparison of records
* Easy maintenance
* Flexible
* Cheap
* Should have some artificial intelligence
* Secure
* Forecasting should be allowed
 * Record processing should be easy

Q. Explain the latest and other important sources of long-

term and short-term financing. (June 02)

Following are some long-term & short-term sources of

financing for a business:

Long-term sources

Shares: It is the most important source for raising permanent

or long-term capital. Section 86 of Companies Act, 1956
provides that share capital of a company formed after April
1, 1956 or the share capital issued after that date, shall be of
only two kinds, viz. Preference share capital and equity share
capital.
Preference shares: According to Section 85 of The
Companies Act, 1956, a preference share is one, which
fulfills the following conditions:

1. A preference share has a preferential right to dividend to

be paid either as a fixed amount or an amount calculated by a
fixed rate which may be either free of or subject to income
tax.
2. A preference share has the right to the repayment of
capital before any thing is paid to equity shareholders on the
winding up of the company.

Equity Shares: According to Section 85 of The Companies

Act, 1956, an equity share is a share which is not a
preference share.

Equity shares entitle to whole of the profits earned by the

company, after a fixed dividend on preference shares that has
been paid by it, are equity shares. Equity shares have no right
to either a fixed dividend or repayment of a pre-determined
amount of capital in the event of winding of the company.

Debentures: When a company desires to borrow a

considerable sum of money for its expansion, it invites the
general public to subscribe to its debentures. A debenture is a
certificate issued by the company acknowledging the debt
due by it to its holders and is issued by means of a prospectus
in the same manner as shares. The following are the various
types of debentures issued by a company: Simple or naked
Debentures, Secured and Mortgage Debentures, Redeemable
Debentures, Perpetual or Irredeemable Debentures,
Convertible Debentures, Non-Convertible Debentures.

Public Deposits: Public deposits are the fixed deposits

accepted by a business enterprise directly from the public.
This source of raising short-term & medium finance was
very popular in absence of banking facilities. Public deposits
have several advantages such as simple & convenient source
of finance taxation benefits, trading on equity, no need of
securities and an inexpensive source of finance.

Ploughing back of profits: It means reinvestment by concern

of its surplus earnings in the business. It is an internal source
of finance & is suitable for an established firm for its
expansion, replacement, etc.

Loans from Financial Institutions: Several financial

institutions like LIC, State Finance Corporation, Industrial
Development bank, etc. also provide loans. This source is
more suitable for medium term demands of working capital.
Interest is charged at fixed rate on these loans.

Short-term Sources
Trade Credit: It is the credit extended by the suppliers of
goods in the normal course of business. It is an important
source of short-term finance. The credit-worthiness of the
firm and the confidence of its suppliers are the main basis of
securing trade-credit.

Advance payment: Some business houses get advances from

their customers and agents. It is a cheap source of finance.

Installment Credit: In this method assets are purchased and

the possession of goods is taken immediately but the
payment is made in installments. Generally, interest is
charged on the unpaid amount.

Commercial Paper: It represents unsecured promissory notes

issued by firms to raise short-term funds. In India only large
companies enjoy high credit rating & can issue commercial
paper to raise short-term funds.
Deferred Income: These are incomes received in advance
before supplying goods or providing service

Using Computers in Inventory Management

Today, the use of computer systems to control inventory is

far more feasible for small business than ever before, both
through the widespread existence of computer services
organizations (listed in the yellow pages of many telephone
directories) and the decreasing cost of micro computers.
Often the justification for such a computer-based system is
enhanced by the fact that company accounting and billing
procedures can also be handled on the computer.

Most computer manufacturers offer free, written information

on the inventory management systems available for their
computers. In addition, computer service companies often
have material readily available describing the use of their
particular computer "software" programs for inventory
management. These companies provide a good source of
information on general descriptions of particular inventory
management techniques, as well as help on specific
inventory management problems.

Whether a manual or computerized inventory management

system is used, the important thing to remember is that
inventory management involves two separate, but closely
related elements: the first is knowing what and how much to
order, when to order and what price to pay; the second is
making sure that the items, once brought into inventory, are
used properly to produce a profit.