Você está na página 1de 216

BUSINESS RESEARCH METHODS

Dr. S. BISALIAH
MARCH, 2013
BUSINESS RESEARCH METHODS
INSTRUCTOR: Dr. S. BISALIAH
1.Donald R Cooper, Pamela S. Chindler and J.
K. Sharma: Business Research Methods 11
th
Edition. The Mcra! "ill #$1#.
#.Deepa% Cha!la and &eena Sondhi: Research
Methodolo'(: Concepts and cases. )i%as
Pu*lishin' "ouse, #$11.
+.,n( other re-erence cited *( the instructor
-rom time to time.
Con-ucius: E.er( truth has -our corners: ,s a
teacher / 'i.e (ou one corner, and it is -or (ou to
-ind the other three.
MODULE 1: Introduction to BR
1.Research and BR:
Pa'e # o- #10
Re:
Research:
1,ll pro'ress is *orn o- in2uir(. Dou*t is
o-ten *etter than o.er con-idence, -or it
leads to in2uir( and in2uir( leads to
in.ention3. "udson Ma4im.
, s(stematic approach to 'ain ne!
%no!led'e.
, s(stematic search -or ans!er to
2uestion or solution to a pro*lem.
BR:
, s(stematic in2uir( that
pro.ides in-ormation to 'uide
mana'ement decisions on
*eha.iour o- emplo(ees, stoc%
holders, customers etc.
Decision 5 ma%ers: To ma4imi6e
*usiness per-ormance.
Pa'e + o- #10
BR: , process o- plannin'
ac2uirin', anal(6in',
disseminatin' rele.ant data 7
in-ormation to help in
decision ma%in' to ma4imi6e
*usiness per-ormace.
BR: &eeded in pro-it and non8pro-it
ma%in' companies.
#.Return on In!est"ent #ROI$: Strate'ies and
tactics to ma4imi6e ma4. R9/. "ence assess
*usiness research e4penditures -or their
contri*ution to R9/.
+.%h& stud& B.R "ethods' %hat sti"u(atin)
*actors::
More .i'orous competition: lo*al and
domestic.
Pa'e ; o- #10
To meet the challen'e o- the *usiness
en.ironment: et 2ualit( data 7 in-ormation
and anal(se it -or emer'in' comple4 decision
5 ma%in' process

need s%ills in BR
methods.
reater computin' po!er and speed


E4plosi.e 'ro!th and in-luence o- the
internet.
&e! perspecti.es on esta*lished Research
Methodolo'ies.
Demand *( sta%e holders in *usiness to
participate in decision ma%in'.
Challen'es o- o.ernment inter.ention in
Corporate Sector.
BR:
o To identi-( and de-ine opportunities and
pro*lems
o To de-ine, monitor, and re-ine strate'ies.
Pa'e < o- #10
o To de-ine, monitor, and re-ine tactics.
o To impro.e our understandin' o- .arious
-ields o- mana'ement.
o Researcher to !or% to!ards a 'oal,
!hether immediate or -uturistic, else the
research loses its si'ni-icance in the -ield
o- mana'ement.
Mana'ement Education: Maturin' as an
independent discipline.
+.Hierarch& o inor"ation , Based Decision
Ma-ers.
..%hat dri!es B.R': /(annin)
Strate'ic and tactical plannin' possi*le !ith
in-ormation 7 data 'enerated *( BR.
Pa'e 0 o- #10
E0hi1it: 1.2. /.13
Strate)&: ,pproach *( the or'ani6ation to
achie.e its 'oal E4. Strate'( -or customer
satis-action, customer lo(alt( etc.
Tactics: BR to help in desi'n and selection
o- tactics to e4ecute a strate'(.
Product Tactic: =hich -la.our in
ice8cream:
Packaging tactic: Container shape
and colour:
Distribution tactic: >ree6er cases or
hand 5 pac%ed drums.
,nd so on.
Re"e"1er "u(ti4(e B , )oa(s: Sales,
mar%et share, return on in.estment,
pro-ita*ilit(, customer satis-action, customer
retention, emplo(ee producti.it(, machine
Pa'e ? o- #10
e--icienc(, ma4. o- stoc% price @or o!ners
e2uit(A and so on.
5.Three "a6or sources o Inor"ation or
Mana)eria( Decision , Ma-in):
BR
Decision support s(stem @DSSA
Business intelli'ence.
DSS: Transactional Data.
Collect data 7 in-ormation at the time o-
transaction o- *usiness E4: Customer data
*ase.
9r'ani6e this in-ormation and retrie.e it
-or decision ma%in'

These data
elements constitute DSS
,nal(sis o- transactional data -acilitated
*( computer technolo'(.
Intranet and E0tranet:
Pa'e B o- #10
Intranet: Pri.ate net !or% !ithin the
compan(
Main purpose: Share compan(
in-ormation !ith internal units -or
decision ma%in'.
E0tranet: ,n e4ternal portion o- a
compan(Cs intranet. Connections to the
outside intranet to share an or'ani6ations
in-ormation !ith e4ternal suppliers,
.endors or customers.
o Because o- DSS, access to date as and
!hen transactions completed. E4:
Customer data *ase to -ind out !hat
tactics lead to a transaction. "o!
pro-ita*le is each customer to the compan(

!ith this strate'ic and tactical plannin'


possi*le.
o Business Inte((i)ence S&ste":
Pa'e D o- #10
9n 5 'oin' in-ormation a*out e.ents and
trends: Technolo'ical, economic,
political, competiti.e arena and so on.

7. Ho8 the Research Industr& 8or-s'
=ho conducts B. R:
E4hi*it ,ppendi4 1.a 5 1, Pa'e #0.
A. Interna( Research Su44(iers:
/nternal Research Department
Small research !in'
/n-ormation technolo'( mana'ement and
research.
ETo pro.ide actiona*le insi'hts, reducin' ris%
in mar%et place actions and impro.in' R9/.
Pa'e 1$ o- #10
E0hi1it: 1 5 1. Pa'e D
B. E0terna( Research Su44(iers:
Business Research >irms:
o >ull8Ser.ice Research >irm
o Research plannin' -or their clients
o E4pertise and -acilities -or *oth
2uantitati.e and 2ualitati.e methodolo'ies
o Cross8>unctional Research Teams.
Custom Researchers
o ,dhoc or custom8desi'ned research
o , research desi'n uni2ue to the decision8
ma%ers dilemma
o ,reas li%e customer satis-action, emplo(ee
moti.ation, product e.aluation
o >ull8Ser.ice Researchers: >it into custom
research cate'or(.
Proprietar( Methodolo'( Researchers:
Pa'e 11 o- #10
o , research techni2ue o!ned *( a sin'le
-irm
E4: The 'allop or'anisation: Sur.e(
methodolo'( uses 1# 2uestions to measure
customer mana'ement. Cop( ri'ht7patent
o- these 2uestions.
Specialit( Research >irms:
o Methodolo'( Specialists:
E4: Sur.e( Research, customer
satis-action, pac%a'in' e.aluation, retail
shoppin' etc.
o Process Specialists: 9nl( one portion o-
the research process.
E4: Telephone inter.ie!in'.
o /ndustr( Specialists: Entertainment
research or tele communication research.
o Participant roups: Children, doctors etc.
o eo'raphic Re'ion Specialists: 9nl( one
re'ion o- a countr( or a 'roup o- countries
li%e South East ,sia.
Pa'e 1# o- #10
S(ndicated Data Pro.iders:
o =ant corporate per-ormance and opinion
data on competin' -irms8sales, mar%et
share, salar( *ene-its etc.
o Then tern to s(ndicated data pro.iders.
o S(ndicate data pro.iders trac%s the chan'e
o.er time in a 'i.en industr( *( -ollo!in'
proper methodolo'(.
o Su*scri*ers has -ull access to its data and
the composite data, *ut not to an
indi.idual competitors data.
9mni Bus Researchers:
o Decision ma%ers need ans!er to one or a
-e! 2uestions li%e pu*lic relation crisis,
product recall or the indictment o- a
compan( e4ecuti.e -or -raud.
Communication ,'encies:
Pa'e 1+ o- #10
o ,'encies to industr( *( placement
e--ecti.eness, e--ecti.eness o- .arious
creati.e approachers @E4: Cele*rit(
endorsin' the compan( productA,
e--ecti.eness o- communication strate'ies
@humour, .iolence etc.A, the R9/ on
.arious media *u(s and on the
comparati.e e--ecti.eness o- di--erent
action stimulants @coups .s samplesA.
Consultants:
o 9--er a !ide ran'e o- ser.ices at the
strate'ic and tactical le.els, usin'
secondar( data research.
o /n-luence o- consultants on research
desi'n, interpretation o- results etc.
o Some consultants conduct *oth
2uantitati.e studies @E4pert opinionA, and
2ualitati.e studies on %no!led'e,
attitudes, opinions and moti.ations.
Trade ,ssociations:
Pa'e 1; o- #10
o To promote, educate and lo**( -or the
interests o- their mem*ers
o Mainl( commission research that ad.ances
trade interests.
9. T&4es o Research Studies:
o MaFor t(pes o- *usiness research
Ma6or T&4es o Business Research
Pa'e 1< o- #10
Basic
Research
,pplied
Research
E4plorator(
Research
Conclusi.e
Research
Descripti.e
Research
Causal
Research
o Mana'ement

Decision8ma%in' !ith
respect to plannin', or'ani6in', controllin',
sta--in', coordinatin' and inno.atin'.

&eed di--erent t(pes o- research support.



o >undamental @PureA R and ,pplied R:
>R @PRA: De.elop *asic scienti-ic
principles
E4: Ps(cholo'ical principles o- "uman
Beha.iour
: Theor( o- Consumer Beha.iour
: Theories o- Moti.ation.
,R: ,pplication o- theor( o- consumer
*eha.iour or theories o- moti.ation to
di--erent areas o- mana'ement li%e "RM,
Mar%etin' Mana'ement @,d.ertisement
response, mar%etin' *eha.iour o-
consumersA.
Pa'e 10 o- #10
o Descripti.e Research: >act >indin' 5 !ho,
!hat, !hen, !here and ho!
Pro*lems o- relationships *et!een
emplo(er and emplo(ee
)ie!in' T) pro'rames *( di--erent t(pes
o- consumers
Descri*e the characteristics o- di--erent
'roups o- consumers -or a particular
product
o Dia'nostic stud(: Concerned !ith casual
relationships !hereas descripti.e stud(
concerned !ith !hat is happenin'.
o >ormulati.e or E4plorator( Research:
o /n.esti'ate a pro*lem li%e lo!
producti.it( o- emplo(ees in a compan(
o T!o su*di.isions o- e4plorator( research:
Pa'e 1? o- #10
Giterati.e Sur.e(: Re.ie! o- =P/,
periodic stoc% e4chan'e inde4 reports,
population trends etc.
E4perience Sur.e(: Sur.e( o-
e4periences o- e4perts in a particular
-ield li%e port-olio mana'ement, product
pricin', technolo'( -orecastin' etc.
o E4perimental Research:
Primaril( in the area o- ph(sical and
*iolo'ical sciences
Possi*le in social sciences li%e
mana'ement also
E4: Simulatin' Consumer Mar%etin'
Beha.iour
: Taste o- a so-t drin% *( di--erent
pro-essionals.
Hse techni2ues li%e ,&9),
Be-ore8a-ter research or *e-ore8a-ter
desi'n !ith control 'roup.
Pa'e 1B o- #10
o "istorical Research:
Stud( o- a mana'ement rele.ant
phenomenon o.er a period o- time 5 state
o- /ndian la*our o.er o- time, Business
Education o.er a period o- time, trends in
/ndian population since 1D$1.
o E48Post8>acto Research:
,n empirical en2uir( -or e.ents that ha.e
alread( occurred
E4: Mar%et -ailure o- a compan(Cs product
: Deceleration o- 'ro!th rate in demand
-or a compan(Cs product.

o Cast Stud( ,pproach:
Micro le.el studies
Stud( o- a *usiness school, a *an%,
per-ormance o- a compan( and so on
>inancial, health o- a compan(.
Pa'e 1D o- #10
o Conclusi.e Research:
De.elop h(pothesis *ased on an
e4plorator( research
Hnder CR, test the h(pothesis and dra!
de-inite conclusions.
o E.aluation Studies:
, t(pe o- applied research
,ssess the e--ecti.eness o- a social or
economic pro'rame
Concurrent e.aluation, periodic e.aluation
and terminal e.aluation.
o ,ction Research:
T(pe o- e.aluation stud(
Concurrent e.aluation stud( o- an action
pro'ramme
Pa'e #$ o- #10
E4: "RD pro'rammes
: Mana'erial impro.ement pro'rammes.
:. Criteria o ;ood Research:
o Purpose and Concepts Clearl( De-ined
o Detailed Research Process:
Descri*e in detail the research procedure
Speci-( the sources o- data and
methodolo'( o- o*tainin' the data to -ind
out the .alidit( and relia*ilit( o- the data
and stud( conclusions
/- no secrec(, some other researchers
should *e a*le to repeat the research.
o Research Desi'n Throu'hl( Planned:
Samplin' desi'n
Minimi6e7eliminate personal *ias
Pa'e #1 o- #10
>ollo! hi'h ethical standards
Re.eal limitations -ran%l(
,nal(sis ade2uate -or decision ma%ers
needs
Present the -indin' unam*i'uousl(
Justi-( conclusions
o ood Research: S(stematic
Go'ical
Empirical
Replica*le
Hse-ul -or decision8ma%in' to
select more e--ecti.e, more
e--icient, less ris% and more
pro-ita*le alternati.e.
13. Research < Scientiic Method:
o BR -or pro*lem sol.in', -ollo!in'
scienti-ic methods
Pa'e ## o- #10
The Essentia( Tenets o the Scientiic
Method:
o Proper statement o- the pro*lem
o 9*Fecti.it(
o De'ree o- ,ccurac(
o Direct o*ser.ation o- phenomenon,
!here.er possi*le
o De-ine clearl( .aria*les, methods and
procedures
o >ormulate empiricall( testa*le h(potheses
o Conclusions *ased on statistical anal(sis.

Reasonin' crucial -or success in research:


Collect data consistent !ith the pro*lem,
-ormulate proper h(potheses, de.elop rele.ant
empirical tests, and deduce and conclude.
T8o Methods o Sound Reasonin):
Induction:
o Stud( se.eral indi.idual cases and dra!
a 'enerali6ation

9*ser.e and 'enerali6e.


Pa'e #+ o- #10
o Con-irm *ased on more e.idence
o Hse this method to ne! -acts
o The conclusion *ecomes onl( a
h(pothesis
E4: =ith considera*le amount o-
e4penditure, the sales o- a compan( do
not increase

=h(: >ormulate additional


h(potheses
o Essential conditions -or satis-actor(
induction:
Correct o*ser.ation and proper
accurate data collection
Select representati.e cases -rom the
stud( uni.erse
,de2uate num*er o- cases
Dra! in-erences -rom the -indin's.
o T!o t(pes o- induction:
Enumerati.e: eneralisation -rom
samples
Pa'e #; o- #10
,nal(tic: Case8*(8case anal(sis.
Deduction:
o Both true and .alid
o Premises @ReasonsA stated in the
conclusion must a'ree !ith the real
!orld @trueA
o Conclusion -ollo!s -rom the premises
E41:
MaFor Premise: ,ll compan(
emplo(ees -ollo! pro-essional ethics
Minor Premise: Krishna is a compan(
emplo(ee
Conclusion: Krishna -ollo!s
pro-essional ethics.

E4#:
MaFor Premise: ,ll permanent
emplo(ees are hard !or%in'
Minor Premise: Rama is a permanent
emplo(ee
Pa'e #< o- #10
Conclusion: Rama is hard !or%in'.
Co"1inin) Induction and
Deduction:
o Hse *oth to8'ether in research reasonin'

1Dou*le mo.ement o- re-lecti.e


thou'ht3 De!e(.
o
Re=uisites o a ;ood Scientiic Method:
o Go'ical anal(sis o- a pro*lem
o De-inition o- concepts and statical units
I measures
o Data collection and classi-ication
o =here possi*le, de-ine .aria*les in
2uantitati.e terms
o Proper statical procedure
o Test h(pothesis and dra! 'eneralisation
and conclusions -rom the stud(.
Co"4onents o Scientiic A44roach:
Pa'e #0 o- #10
E4hi*it +.B I +.D PP: B+ I B;
o De-ine the pro*lem
o >ormulate h(pothesis as to
causes7e4planations7solutions o- the
pro*lem
o Collect data
o ,nal(se the data, test the h(pothesis and
dra! in-erences.
Conce4ts: Meanin) and Sources
o enerall( accepted collection o-
meanin's or characteristics associated
!ith certain e.ents, o*Fects, conditions,
situations and *eha.iours.
E4: spreadsheet, !arrant( card
E4: !al%in', runnin'.
o Sources:
9ther lan'ua'es
9ther disciplines
Pa'e #? o- #10
E4: ra.itation used in mar%etin' to
e4plain !h( people shop !here the(
do.
Distance used in attitude
measurement.
&e! meanin's -or concepts:
E4: Model.
o /n research, need -or concepts precision
and in.enti.eness
E4: Respondents to estimate their -amil(
income.
But Precision: Time period, *e-ore or
a-ter income ta4, head o- -amil( or all
mem*ers, income in %ind such as -ood
stamps, -ree rent etc.
o Constructs:
,*stract ideas used in a 'i.en
research.
Build constructs *( com*inin' the
simpler, more concrete concepts !hen
Pa'e #B o- #10
the idea or ima'e not su*Fect to direct
o*ser.ation.
E4: Ta*le: ,n o*Fecti.e concept
Personalit(: Much more di--icult
to .isuali6e
Satis-ied ser.ice customer:
o Concept Mappin':
, -ormati.e assessment and a research
tool
, techni2ue -or .isuali6in' the
relationship *et!een concepts
, dia'ram sho!in' the relationship
*et!een concepts
To communicate comple4 ideas.
11. >aria1(es:
o , s(m*ol o- a trait or attri*ute that can
*e measured.
o Dichotomous )aria*les: T!o .alues
presence or a*sence o- a propert(
E4: Emplo(ed 5 Hnemplo(ed
Pa'e #D o- #10
Male 5 >emale.
E4: ,utomoti.e )aria*le
,m*assador car: 0
"onda: ?
&o option -or 0.<
o Continuous )aria*les: /ncome, a'e,
temperature.
o /ndependent and Dependent )aria*les:
Does ad.ertisement in-luence product
purchase
/ndependent )aria*le:
Dependent )aria*le:
More than one independent .aria*le:
Product PurchaseJ @,E, ,'e,
Giterac( Ge.el, /ncome KK.A
o Moderatin' )aria*le:
Second independent .aria*le *elie.ed
to ha.e si'ni-icant contri*utor( e--ect
on the dep. H.
Pa'e +$ o- #10
o E4traneous )aria*les:
Some .aria*les independent or
moderatin' .aria*les
But man( .aria*les assumed or
e4cluded -rom the stud(
E.): Gittle or no e--ect on a 'i.en
situation.
o /nter.enin' )aria*les:
Theoreticall( a--ect the dependent
.aria*le, *ut cannot *e seen or
measured.
12. /ro4ositions and H&4otheses:
Proposition: , statement a*out o*ser.a*le
phenomenon
Pa'e +1 o- #10
"(potheses: , proposition -ormulated -or
empirical testin'.
E4: Mana'er L has hi'her8than8a.era'e
achie.ement moti.ation.
Moti.ation is the .aria*le: , Characteristic
or attri*ute.
T(pes o- "(pothesis:
o Descripti.e: B$M o- compan( stoc%
holders @casesA -a.our increasin'
compan(Cs cash di.idend @.aria*leA.
E Research Nuestion: =hat percenta'e
o- compan( shareholders -a.our cash
di.idend:

Descripti.e h(pothesis is use-ul -or


testin' statistical si'ni-icance.
o Relational "(potheses: Descri*e a
relationship *et!een t!o .aria*les !ith
respect to some case.
Pa'e +# o- #10
E4: >orei'n @.aria*leA MB,s percei.ed
*( /ndians @caseA to *e o- *etter 2ualit(
@.aria*leA than domestic MB,s.
o Correlational "(potheses: )aria*les
occur to'ether in some speci-ied manner
!ithout impl(in' one causes another.
E4: Students o- , Section 'i.e more
-a.oura*le ratin' -or Pro-. L than do
students o- B Section.
o E4planator( @causalA "(potheses:
Chan'e in one .aria*le leads to a
chan'e in the other .aria*le
Causal .aria*le, called independent
.aria*le, and the other dependent
.aria*le
E4: ,n increase in -amil( income @/)A
leads to an increase in the percenta'e
o- income sa.ed @D)A.
The Ro(e o the H&4othesis:
o uides the direction o- the stud(
Pa'e ++ o- #10
o /denti-ies rele.ant -acts
o Su''ests the appropriate research desi'n
o Pro.ides -rame!or% -or or'ani6in' the
conclusion
1?. Theor&:
, set o- s(stematicall( interrelated concepts,
de-initions and propositions
Su''est a pro*lem -or stud(
Su''est a h(pothesis -or testin'
9- help in e4plainin' and predictin' a
phenomenon.
E4: >our sta'es 'enerali6ed o- Product Gi-e
C(cle in Mar%etin'
/n each sta'e, man( concepts, constructs and
h(potheses descri*le the in-luences that
chan'e re.enue and pro-it.
Pa'e +; o- #10
E4hi*it +.< P. ?0
E4: Earl( sta'e, hea.( e4penditure to create
product a!areness
"i'h pricin'

s%immin' concept
Go! pricin'

penetration pricin' concept.


Mode(:
o Theor( to e4plain and Model to represent
o M: , representation o- a s(stem
constructed to stud(
o Predicti.e Model: >orecast -uture e.ents
o &ormati.e Model: To in-orm a*out !hat
actions to *e ta%en sa( to sustain
consumer *rand lo(alt(
o Static I D(namic Model:
Static Model: Represent a s(stem at one
point in time
D(namic Model: E.olution o- a s(stem
o.ertime.
1+. So"e Research /ro1(e"s in Mana)e"ent:
Pa'e +< o- #10
Mar-etin):
o Product identi-ication
o Demand estimation
o Mar%et se'mentation
o Media mi4
o Product launchin'
o Distri*ution model desi'n
o Pac%a'in'
o Bu(in' *eha.iour
o ,nd so on.
/roduction:
o >orecastin'
o Plant location
o /n.entor( control
o Production schedulin'
o Nualit( control
o ,nd so on.
*inance:
o Brea%8e.en anal(sis o- a sin'le product
and multi8product
Pa'e +0 o- #10
o Capital *ud'etin'
o Ratio anal(sis
o Port-olio mana'ement
o ,nd so on.
HRM #/ersonne( Mana)e"ent$:
o Manpo!er plannin'
o Per-ormance appraisal s(stems
o Con-lict mana'ement
o Trainin' methods
o Chan'e mana'ement
o ,nd so on.
Pa'e +? o- #10
MODULE 2
RESEARCH /ROCESS
1. Deine Ste4s in Research:
2. Identi& Mana)e"ent , Research @uestion
Hierarch&:
State the *asic mana'ement dilemma,
leadin' to the need -or research
Brea%do!n the *asic 2uestion into more
speci-ic ones

Mana'ement 5 Research 2uestion


hierarch(
Mana'ement Dilemma: S(mptom o- a
Mana'ement Pro*lem
o Risin' product7ser.ice cost
Pa'e +B o- #10
E4hi*it ;.1 P. D1
E4hi*it ;.# P. D+
o /ncreasin' tenant mo.e8outs -rom an
apartment comple4
o Declinin' sales
o /ncreasin' emplo(ee turno.er in a
restaurant
o , lar'e num*er o- product de-ects durin'
the manu-acture o- an automo*ile
o ,n increasin' num*er o- letters and phone
complaints a*out post purchase ser.ice
o ,n( other delimma.
Mo.e -rom Mana'ement Dilemma to
Mana'ement Nuestion: Restate the dilemma
in 2uestion -orm
o =hat should *e done to reduce emplo(ee
turno.er:
o =hat should *e done to increase tenant
residenc( and reduce mo.e8outs:
o =hat should *e done to reduce costs:
o ,nd so on.
Pa'e +D o- #10
Three cate)ories o Mana)e"ent
@uestions:
o 9*Fecti.e: =hat do !e !ant to achie.e:
E4: Dou*le sales and pro-its in < (ears.
"o! can !e achie.e the o*Fecti.e setout:
E4: =hat should *e done to reduce post
purchase ser.ice complaints in case o-
Mind!riter e4ample.
: "o! can !e achie.e -i.e (ears 'oal o-
dou*led sales and net pro-its:
o Trou*le shootin' or control situation:
Monitor .arious !a(s in !hich an
or'anisation is succeedin' or -ailin' to
achie.e its 'oal.
?. The Nature o the Mana)e"ent @uestion:
An E0a"4(e
Pa'e ;$ o- #10
, researcher to help the mana'ement o- a
*an% -acin' erosion o- the *an%Cs
pro-ita*ilit( @the mana'ement dilemmaA
"o! can the pro-it picture *e impro.ed:

The mana'ement 2uestion is *road.


T!o speci-ic 2uestions *ased on discussions
*et!een the researcher and the mana'ement
o "o! can the *an% increase deposits:
o "o! can the costs o- *an% operations *e
reduced:
+. E04(oration and *or"u(ation o Research
@uestions:
Mo.e -rom mana'ement 2uestions to
research 2uestions: Recall "ierarch(
o Research Nuestions:
Pa'e ;1 o- #10
=hat contri*utor( -actors -or lo!
'ro!th rate in deposits:
/n the Ban%: "o! is !or% climate,
e--icienc( o- operations compared to
Ban% /ndustr( &orms, and -inancial
conditions compared to industr( norms
and competitors:
o Conduct -ocus 'roup discussions !ith
emplo(ees, 'et data on -inancial and
operatin' statistics -rom the Ban% and
annual and other reports.
o Conduct literature search.

Hnstructured e4ploration E De.elop and re.ise


mana'ement 2uestions E Translate
these into research 2uestions.
Research Nuestion: , -act8oriented and
in-ormation8'atherin' 2uestion
Researcher to discuss !ith the mana'ement
thorou'hl(, *e-ore he -ormulates the
research 2uestion a'reea*le to the
mana'ement
Pa'e ;# o- #10
>ine8Tunin' Research Nuestion:
o E4amine the concepts and constructs
rele.ant -or the stud(
o Re.ie! research 2uestions and *rea% them
do!n into speci-ic second8and third le.el
2uestions
o "(potheses -ormulation and tests to *e
per-ormed
o Set the scope o- the stud( state !hat is not
a part o- the research 2uestion.
+.1 In!esti)ati!e @uestions:
o Continue e4plorator( stud(, re-ine the
ori'inal 2uestions, and construct
in.esti'ati.e 2uestions.
o /n.esti'ati.e 2uestions 'uide the research
desi'n.
Pa'e ;+ o- #10
o The e4ample o- lac% o- deposit 'ro!th in
the stud( Ban%:
=hat is the pu*licCs position re'ardin'
-inancial ser.ices and their use:
Three su*2uestions @P.00A
=hat is the Ban%Cs competiti.e
position:
Si4 su*2uestions @P.00A.
+.2 Research /rocess Issues:
Hse the -a.oured8techni2ue
E4: Sur.e(7case stud(
E4: TNM: Standarised customer satis-action
2uestionnaire a.aila*le.
o Research Techni2ues @Details later ch: ?8
11 o- Cooper and PamelaA.
+.? Co"4an& Data1ase Minin):
M/S o- the compan(
Pa'e ;; o- #10
>ind out !hat data a.aila*le, *e-ore startin'
data collection

Data*ase minin': , startin' point.


+.+ Desi)nin) Research Stud&:
R. Desi'n: Blueprint -or meetin' the
o*Fecti.es o- the stud( and ans!erin'
research7mana'ement 2uestions.
Ma%in' choice -rom lar'e .ariet( o-
methods, techni2ues, procedures and
samplin' plans
E4 : Secondar( data stud(, case stud(,
sur.e(, e4periment or simulation
: 9- sur.e(, mailin' the 2uestionnaire,
computer, telephone or the personal
inter.ie!
: =hat %ind o- 2uestionnaire, !hat
2uestions:
: Response to *e scaled or open ended
Pa'e ;< o- #10
: "o! to ensure relia*ilit( and .alidit(:
: =hat %ind o- trainin' to data collector:
: Sample: census or samplin':
Sa"4(in) Desi)n:
o /denti-( the tar'et population
o Select the sample7census
o "o! man( samples:
o Samplin' techni2ues: !hich one: @ch 1;A.
Research Bud)et and Resource
A((ocation: Ho8 "uch or:
o Research proFect plannin'
o Data collection
o ,nal(sis and interpretation
Pa'e ;0 o- #10
o Report !ritin'
E E2ual or une2ual share in total *ud'et.
o Bud'etin': Criteria
Some M o- sales re.enue
Departmental or -unctional area
*ud'etin'
E4: Some M o- *ud'et o- "RM or
Mar%etin' Department
Tas% Bud'etin': ,d hoc allocation -or
a speci-ic R8proFect.
o Bud'et *ased on: )alue o- research
in-ormation.
o Research result: E4 to produce added
re.enue or reduce costs

Research Bud'et: ,n in.estment.


E!a(uation o >a(ue o Research:
Pa'e ;? o- #10
o E4post >acto E.aluation: Measure the
contri*ution o- research to corporate
pro-ita*ilit( a-ter8the8-act e.ent E.
Measurement o- return8on8in.estment
a-ter the e.ent E. Can per-orm cost8
*ene-it anal(sis.
o 9ption7choice ,nal(sis and Decision
Theor(:
Methodolo'ies -or estimatin' costs
and *ene-its o- R 5 stud(
Costs: Relati.el( eas( to estimate,
*ene-its di--icult to estimate
Turn to Decision Theor(: =hich
criterion to decide: Com*ination o- a
decision rule and a decision .aria*le.
E4 : Decision )aria*le: Contri*ution
to pro-its or resource sa.in' e--ect o-
research proFect.
Pa'e ;B o- #10
: Decision )aria*le: E4pressed in
monetar( terms: Sales re.enue, costs,
pro-its

Nuanti-ia*le measure.
E4 : Decision Rule: Choose the
alternati.e that pro.ides the 'reatest
annual net pro-it.
: Mind =riter: Choose the alternati.e
that pro.ides the hi'hest le.el o- post
ser.ice satis-action.
+.. The Research /ro4osa(:
Contents o Research /ro6ect /ro4osa(:
T8o Basic Sections
Statement o- the research 2uestions recall
-rom Ban% ProFect.
Brie- description o- research
methodolo'(7Research Desi'n.
Pa'e ;D o- #10
E4hi*it ;.+ P. D0
/i(ot Testin): An i"4ortant sta)e in dataA
)atherin) 4hase
o To detect !ea%ness in research desi'n and
research instrument.
Data Co((ection: /art III o Coo4er et. a(
o Data collection methods
o )eri-ication o- data7consistenc(
o Data editin'
o Codin'
o Data entr(.
Ana(&sis and Inter4retation: /art I> o
Coo4er et. a(
Reduce collected data to a mana'ea*le si6e
De.elop summaries
o ,ppl( statistical techni2ues
Pa'e <$ o- #10
o /nterpretation, %eepin' in .ie! clients
research 2uestion
o Ma%e recommendations.
E4: Mar%et research on pre-erence -or
!allet 5 si6ed porta*le telephone .is8O8.is
cellular telephones.
Re4ortin) Research Resu(ts:
9r'anisation and st(le o- report, %eepin' in
.ie! the clients needs7mana'ers needs -or
decision8ma%in'.
Care-ul choice o- !ords in cra-tin'
interpretations, conclusions and
recommendations.
Contents o the research re4ort:
o ,n e4ecuti.e summar(: S(nopsis o- the
pro*lem -indin's and recommendations
Pa'e <1 o- #10
o ,n o.er.ie! o- the research pro*lems
*ac%'round, literature summar(, methods
and procedures and conclusions
o /mplementation strate'ies -or the
recommendations
o Technical ,ppendi4: ,ll the materials
necessar( to replicate the research proFect.
+.5 A S&no4tic Re4resentation o Research
/rocess:
De-ine Research Pro*lem
Re.ie! Giterature
@Concepts and theories
Pre.ious research >indin'sA
>ormulate "(potheses
Research Desi'n
Collect Data
Pa'e <# o- #10
@E4ecutionA
,nal(se Data
@Test "(potheses i- an(A
/nterpret and Report.
E Can *e a -eed*ac% -rom data
anal(sis to data collection.
E Can *e a -eed*ac% -rom research
desi'n to data anal(sis and .ice8.ersa.
E >eed*ac% -rom interpretation and
report !ritin' to data anal(sis.
Pa'e <+ o- #10
MODULE ?
BUSINESS RESEARCH: RE@UEST AND
/RO/OSAL
1. The re=uest or R. /ro4osa(:
, -ormal document *( a corporate research
department
Creatin' the Re2uest -or Proposal: R>P
o R>P: /nternal e4perts to de-ine the
pro*lem
E4: Brand Mana'ers &e! Product
Specialist
o Bidders to recei.e the R>P:
o The components o- R>P:
/n-ormation on the administration o-
the proFect
Pa'e <; o- #10
Summar( o- statement o- the pro*lem
con-rontin' the compan(
Technical Section:
Pro*lem statement
Description o- -unctional re2uirements
/denti-ication o- constraints.
Mana'ement Section: Timin' on
schedules, plans and reports.
Contracts and Gicense Section:
T(pe o- contracts to *e si'ned *( the
*idder
E4: &on8disclosure o- research
in-ormation.
Pricin' Section:
Proposal Costin': components.
*or"at o R*/:
Pa'e << o- #10
o /nstructions to *idders
o ProFect *ac% 'round
o )endor in-ormation:
E4: Biddin' compan( pro-ile
o Proposed solution
o Ser.ices and support
o Cost proposal.
2. The Research /ro4osa(:
,n indi.idualCs or compan(Cs o--er to render
the ser.ice7produce a product
Purpose o- the research proposal
o Research on mana'ement 2uestions
o Discuss the earlier research e--orts on the
2uestions
o Speci-( data needs, data collection, data
anal(sis and interpretations
Pa'e <0 o- #10
o ResearcherCs credentials, plans and
ser.ices so as to present a !innin'
proposal
o Speci-( the *ene-its o- proposed research
to the sponsorin' compan(.
E.aluation o- the proposal *( the sponsorin'
compan(:
o Plannin' o- the research !or%
o Properl( !ritten document in a !ell
or'ani6ed manner
o Clarit( o- research desi'n
o ResearcherCs capa*ilities to underta%e the
proFect
o Mana'ement 2uestions translated into
research 2uestions.

,ssess the proposal and the researcher to


compare a'ainst the competin' proposals.
Pa'e <? o- #10
MODULE +
THE BUSINESS RESEARCH DESI;N
#COO/ER et. a( /ART II: CH.5$
1. Research Desi)n:
Blueprint -or the collection, measurement
and anal(sis o- data
"elps in allocation o- resources *( ma%in'
proper choices in methodolo'(
Plan in.esti'ation structure to o*tain
ans!ers to research 2uestions
E4presses *oth the structure o- research
pro*lem and plan o- in.esti'ation.
2. Research Desi)n:
?. Essentia(s o Research Desi)n:
Pa'e <B o- #10
E4hi*it 0.1 P.10$
o Time8*ased plan
o Plan *ased on research 2uestion
o uides selection o- sources I t(pes o-
data7in-ormation
o Speci-( relationships amon' the stud(
.aria*les
o , procedural outline.
+. C(assiication o Research Desi)ns: Use o
Descri4tors
.. T&4es o Research Desi)n:
E4plorator(
Descripti.e
Causal
,nd so on
Pa'e <D o- #10
E4hi*it 0.+ P.10#
5. So"e E(a1orations:
E4plorator( Studies:
o Goose structure !ith the o*Fecti.e o-
disco.erin' -uture research studies
o The -ormal stud( *e'ins !hen e4plorator(
stud( ends.
Monitorin' and communication Studies:
o Monitorin' Stud(: Monitor the acti.ities
!ithout elicitin' response -rom an( *od(
E4: Tra--ic counts at an intersection
o Communication Stud(:
E4: /nter.ie!7telephone7mailed
2uestionnaire

Method o- Data Collection


o E04eri"enta( and E04ost *acto
Desi)nBStud&:
Pa'e 0$ o- #10
E4perimental: Po!er-ul method -or
testin' h(pothesis causation.


ResearcherCs control o.er .aria*les.
E4post >acto Desi'n:
&o in-luence on the .aria*les and
e.ent
Stud( !hat has happened.

Researcher Control o- )aria*les


o Descri4ti!e and Causa( Desi)nBStud&:
Descripti.e Stud(: >ind out !ho,
!hat, !here, !hen or ho! much
Causal Stud(: >ind out !h( o-
happenin's7e.ents.

The purpose o- the stud(


o CrossASection and Lon)itudina( Stud&:
Pa'e 01 o- #10
CSS: Stud( at a point in
time7once7snapshot o- one point in time.
GS: Repeat the same stud( o.er a
period o- time to trac% chan'es o.er
time.

The time dimension


o Statistica( and Case Studies:
SS: Gar'er *readth o- stud(, more
sample to test h(potheses statisticall(
CS : More emphasis on a -ull
conte4tual anal(sis
: Rel( on 2ualitati.e data

Di--icult to test h(potheses


: )alua*le insi'hts -or pro*lem
sol.in' o- a compan(.
Research En.ironment:
o Studies under actual7en.ironmental7-ield
conditions
o Studies under la*orator( conditions
Pa'e 0# o- #10

Simulation studies.
/artici4ants /erce4tions o Research
Stud&:
o /- the emplo(ees percei.e that some stud(
is *ein' conducted on their 1Moti.ation3,
li%el( less than the natural *eha.iour.
o Ma( in-luence the -indin's o- the stud(.

&eed -or .i'ilance on the part o- the


researcher
7. E04(orator& Studies:
o E4ploration : Researchers lac% a clear idea
o- the research pro*lems
: To de.elop concepts, esta*lish
priorities, de.elop operational
de-initions, and to impro.e
research desi'n.
: Sa.e mone( and time
Pa'e 0+ o- #10
: /denti-( .aria*les and -ormulate
h(potheses.
o E4ploration: Relies more hea.il( on
2ualitati.e techni2ues, instead o- on
2uantitati.e techni2ues.
@ua(itati!e Techni=ues:
o Nualit(: Essential character or nature o-
somethin'
o Nuantit(: The amount
o Nualit( is the !hatP Nuantit(: The ho!
much
o Nualitati.e: Meanin'7de-inition o-
somethin', and 2uantitati.e: Measurement
o- it.
o Se!era( a44roaches in @ua(itati!e
E04(orator& Studies:
#Detai(s in ch. 5$
Pa'e 0; o- #10
Depth inter.ie!s o- indi.iduals
Participant o*ser.ation
>ilms7Photo'raphs7)ideotape
ProFecti.e techni2ues and
ps(cholo'ical testin':
E4: Role pla(in'
Case studies
Street ethno'raph(:
E4: Cultural 'roup at the street le.el
E4pert inter.ie!in'
Document anal(sis

*our E04(orator& Techni=ues:


Secondar( data anal(sis
E4perience sur.e(s
>ocus 'roup
T!o8sta'e desi'ns.
Pa'e 0< o- #10
Secondar& Data Ana(&sis:
o Search o- the secondar( literate @i.e studies
*( others represent secondar( dataA

/ne--icienc( o- e--orts to collect primar(


data7to conduct ori'inal stud(, i- su--icient
in-ormation a.aila*le in other studies -or
mana'erial decision8ma%in'.
o Sources o- secondar( data:
Search in an or'anisations o!n data
archi.es
Pu*lished documents o- other
or'anisations, a.aila*le sa( in an
internet search
E04erience Sur!e&:
o See% in-ormation -rom persons
e4perienced in the stud( area
o /nter.ie! the e4perienced persons:
Pa'e 00 o- #10
o To de.elop ne! h(pothesis, i- !arranted
and to discard the old one.
>ocus roups:
o , 'roup o- people led *( a trained
moderator
o Moderator @>amiliar !ith 'roup
d(namicsA: uide the 'roup in an
e4chan'e o- ideas, -eelin's and
e4periences on a speci-ic topic.
E4: &e! product7product concept7a ne!
emplo(ee moti.ation pro'ram
E4: Consumer studies.

The *asic output o- the session: /deas,


*eha.ioural o*ser.ations

Enrich research
2uestions7h(potheses -ormulation.
T8o Sta)e Desi)n in E04(orator& Studies:
Pa'e 0? o- #10
o >irst Sta'e: To de-ine the research
2uestion and to de.elop the research
desi'n.

To identi-( the maFor dimensions o- the


research !or%

De-ine su*sidiar( in.esti'ati.e 2uestions

De.elop h(potheses a*out the possi*le


causes o- a mana'ement dilemma.
9. Descri4ti!e Studies:
&ot an e4plorator( stud(, *ut a -ormali6ed
stud( !ith clearl( de-ined
h(potheses7in.esti'ati.e 2uestions
Descripti.e studies to ser.e research
o*Fecti.es:
E4 : Description o- characteristics o-
a population @sa( tar'et populationA
Pa'e 0B o- #10
: =ho, !hat, !hen, !here and ho! o-
the topic
: Estimate proportion o- population
!ith these characteristics
: Disco.er associations amon'
.aria*les @correlation stud(A

Correlational Stud(: , su*set o- descripti.e


stud(.

Demandin' o- research s%ills as in the


causal stud(.
E4 : Recall Ban% Stud(: De.elop a pro-ile o-
sa.ers, relati.e si6e o- accounts, minor
accounts, num*er o- deposits7!ithdra!als
and so on.
: Conduct a sample sur.e( o- sa.ers: >amil(
li-e c(cle, attitude to8!ards sa.in's, -amil(
income le.els and so on.

Correlate this in-ormation !ith data on


sa.in's.

Data7in-ormation collected leadin' to causal


2uestions. E4: =h( !ould people !ho li.e
Pa'e 0D o- #10
-ar a!a( -rom the Ban% o--ice ha.e an
account in the Ban%.
:. Causa( Studies:
Correlation *et!een location and pro*a*ilit(
o- account holdin' in the Ban%.
But correlation is not causation

,ssociation @correlationA does not mean


causation.
The essential clement o- causation:
, produces7-orces B to occur
E4 : Sales -ail to increase -ollo!in' a
sales promotion
: Pro*a*le: Cause: , poorl( e4ecuted sales
promotion

Di--icult to esta*lish cause8e--ect


relationships.
Causa( Re(ationshi4s:
Pa'e ?$ o- #10
o "o! one .aria*le a--ects, or is
1responsi*le3, -or chan'es in another
.aria*le
o /n Business Research, cause8e--ect
relationships less e4plicit
o /n BR, more interest in understandin',
e4plainin', predictin' and controllin'
relationships *et!een .aria*les
o Three 4ossi1(e re(ationshi4s 1et8een
t8o !aria1(es:
S(mmetrical: T!o .aria*les -luctuate
to'ether. But the( ma( *e indicator o-
another cause or /).
E4: Go! !or% attendance and acti.e
participation in a compan( campin'
clu*: Correlated *ut ma( *e due to
another -actor, li-e st(le pre-erence.
Reciprocal Relationship: T!o
.aria*les mutuall( in-luence or rein-orce
each other.
Pa'e ?1 o- #10
E4 : Readin' o- all ad.ertisement


use o- a *rand product.
: Hsa'e

sensiti6es the consumer to


read more o- the ad.ertisin' that
particular *rand.
Ho8 to test Causa( H&4otheses:
o , causes B to occur

B.
o But e.idences on ,

B: Three
Co8.ariation *et!een , I B:
Do , I B occur to'ether in the !a(
h(pothesi6ed
=hen , does not occur, B also does
not occur:
More or less o- ,

More or less o- B
also:
Time order o- e.ents
Does , occur *e-ore B.
Pa'e ?# o- #10
&o other possi*le causes o- B:
Determine that C, D and E do not co8
.ar( !ith B to su''est that there is a
causal relationship *et!een , I B.
Causation and E04ost *acto Desi)n:
o Stud( su*Fects !ho ha.e *een e4posed to
the independent -actor @E4post >actoA and
those !ho ha.e not.
E4: >act: Se.eral !or%ers in a
compan( are a*sent on Monda(sC
Disco.er: Some o- these !or%ers:
Mem*ers o- a campin' clu*
"(potheses: Clu* mem*ership


,*senteeism
Best ,pproach: Chec% the attendance
o- clu* mem*ers: >or a period o- time
and o- those !ho are not mem*ers.
Pa'e ?+ o- #10
MODULE .
DATA COLLECTION: METHODS <
INSTRUMENTS
1. a. Data Source
Data Source
1. T8o T&4es: /ri"ar& and Secondar&
Data
Pa'e ?; o- #10
Primar( Secondar(
/nternal E4ternal
Electronic
Media
S(ndicated
Sources
Pu*lished >ull(
Process
&eed
>urther
Processin'
Data: >acts and other rele.ant materials
: Determine 2ualit( o- the -indin's o- the
stud(
: Basis -or testin' the h(potheses

The scienti-ic process o- measurements,


anal(sis, testin' and in-erences depend on
data.
: T(pes o- data related to human *ein's:
Demo'raphic and socio8economic
characteristics and *eha.ioural .aria*les
@opinions, attitudes, %no!led'e,
a!areness etcA.
T8o Sources #T&4es$ o Data:
o Primar( Data: 9ri'inal data collected
a-resh *( the researcher.
E4: Brand a!areness, *rand pre-erence,
*rand lo(alt( and so on.
o Secondar& Data:
Pa'e ?< o- #10
,lread( collected7compiled and
passed throu'h the statistical process
E4: Census reports, -inancial statements
o- companies, o.ernment Department
reports, &SS data and so on.
Readil( a.aila*le
Researcher: &o ori'inal control o.er
collection and classi-ication
Hses o- S 5 D:
Re-erence purpose
E4: )olume o- *usiness turn8o.er
,s *ench mar%s
Sole source o- in-ormation -or a
research proFect
E4: Securities Mar%et Beha.iour,
>inancial ,nal(sis o- Companies and
so on.
Ad!anta)es o S.D:
Pa'e ?0 o- #10
Secure 2uic%l( and cheapl(
=ider 'eo'raphical area and lon'er
re-erence period
Data -rom se.eral en.ironmental and
social settin's: use-ul -or
'eneralisation
To .eri-( the -indin's *ased on
primar( data.
Disad!anta)es o S.D:
Ma( not meet the speci-ic needs,
de-initions, and measurements.
/naccurac(.
E!a(uation o Secondar& Data:
,.aila*le data rele.ant -or the stud(
on hand: De-initions7 classi-ications
emplo(ed: Measurement o- .aria*les:
Co.era'e:
Pa'e ?? o- #10
Data Nualit(: ,ccurac(, relia*ilit( and
completeness.
T&4es o Secondar& DataBInor"ation
Sources:
o Bi*lio'raphies: 9nline catalo' in the
li*rar(
o Dictionar( o- Business and Mana'ement
o Enc(clopedias: The Enc(clopedia o-
compan( histories
o "and*oo%s: "and*oo%s o- statistics
@Recall RB/ "and*oo%A.
2. Methods o Co((ectin) /ri"ar& Data:
9*ser.ation
/nter.ie!in'
Mail Sur.e(
Pa'e ?B o- #10
E4perimentation
Simulation
ProFecti.e Techni2ue.
?.1 O1ser!ation:
, classical method o- scienti-ic en2uir(. E4:
&atural and ph(sical sciences such as
*iolo'( and astronom(.
, scienti-ic in2uir(, !hen it is conducted to
ans!er a research 2uestion.
=hen it pro.ides relia*le and .alid account
o- the o*ser.ed phenomena.
/ncludes *oth Qseein'C and Qhearin'C.
T&4es o O1ser!ation:
o Beha.ioural:
&on8.er*al anal(sis
Pa'e ?D o- #10
Gin'uistic anal(sis
E4tra lin'uistic anal(sis
Spatial anal(sis.
o &on8*eha.ioural:
Record anal(sis
Ph(sical condition anal(sis
Ph(sical process anal(sis
NonABeha!ioura( O1ser!ation:
o Record ,nal(sis: Common method o-
o*ser.ation o- historical or current records
and pu*lic or pri.ate records
E4: >inancial records, !ritten, sound8
recorded, photo'raphed or .ideotaped.
o Ph(sical Condition ,nal(sis:
,nal(sis o- in.entor( condition
Pa'e B$ o- #10
,udits o- merchandise a.aila*ilit(
o Ph(sical Process ,nal(sis:
E4: Time studies o- manu-acturin' process
: ,nal(sis o- tra--ic -lo!s.
Beha!ioura( O1ser!ation
o Non , >er1a( Ana(&sis:
E4: Bod( mo.ement: Mo.ements o-
!or%er assem*lin' a product
E4: Time spent *( !or%ers on .arious
acti.ities
E4: Motor E4pressions: Si'n o- emotional
status
E4: E4chan'e o- 'lances in interpersonal
*eha.iour studies.
o Lin)uistic Beha!iour:
E4: =ords a pro-essor uses durin' a class
Pa'e B1 o- #10
: Stud( o- !hat, ho!, and ho! much o-
in-ormation con.e(ed in a trainin'
situation
: /nteraction processes
o E0tra Lin)uistic:
E4: )ocal: Goudness
: Temporate: Rate o- spea%in'
: /nteraction: Tendencies to interrupt
and dominate
: )er*al st(listic: .oca*ular( and
pronunciation peculiarities.
o S4atia( Re(ationshi4s:
E4: Stud( o- the e--ects o- cro!din' in a
!or% place.
o Ad!anta)es o O1ser!ation Method:
Collect the ori'inal data o- the time
the( occur. E4: 9*ser.ation o- *u(in'
acti.it( in a store.
Pa'e B# o- #10
Capture in-ormation in a natural
en.ironment
9*ser.ation: Better than elicitin'
response.
o Disad!anta)es o O1ser!ation Method:
Researcher a.aila*ilit( at the time o-
e.ent
Slo! and e4pensi.e
Gi%el( su*Fecti.e assessment
T&4es o O1ser!ation:
o Participant 9*ser.ation:
=idel( used in anthropolo'ical stud(
The o*ser.er is a part o- the
phenomenon
E4: "e is a customer in a retail outlet.
Pa'e B+ o- #10
o &on8Participant 9*ser.ation: Hn%no!n to
people he is o*ser.in' @Dis'uised
o*ser.ationA
o Controlled and Hncontrolled 9*ser.ation:
C9: Standardisation o- o*ser.ational
techni2ues and s(stematic recordin' o-
data. E4: E4perimentation.
HC9: Hsed primaril( -or descripti.e
research
: Hsed to 'et spontaneous picture o-
the e.ent.
o Direct and /ndirect 9*ser.ation:
D9: The o*ser.er, ph(sicall( present
/9: Mechanical, photo'raphic or
electronic recordin'.
O1ser!ation Too(s and Recordin)
De!ices:
Pa'e B; o- #10
o 9*ser.ation schedule:
o >ield o*ser.in' lo': Diar( or Cards
o Mechanical de.ices: Cameras, tape
recorders, .ideotape and so on.
A e8 ti4s or eecti!e use o O1ser!ation
Method:
o Speci-( contents: =hat .aria*les to
o*ser.e
o 9*ser.er Trainin':
Concentration
,*ilit( to remem*er details o- an
e4perience.
o ,*ilit( to e4tract the most -rom
o*ser.ation
o Decide care-ull( tar'et 'roup.
Pa'e B< o- #10
?.2 Inter!ie8in) Method:
S(stematic con.ersation *et!een the
researcher and the in-ormant
>ace8to8-ace or contact o.er telephone
>acilitates clari-ication -rom the respondent
or in-ormant
T&4es o Inter!ie8:
o Personal /nter.ie!7Telephonic /nter.ie!
o Structured /nter.ie!:
Response on a predetermined 2uestions.
o Hnstructured /nter.ie!:
&ot to -ollo! a s(stem o- predetermined
2uestions.
o >ocused /nter.ie!:
>ocus attention on the 'i.en e4perience o-
the respondent.
Pa'e B0 o- #10
E4: )ie!s on a 'i.en -ilm.
o Clinical /nter.ie!:
Encoura'e the respondent to tal% a*out a
'i.en topic
E4: Ps(chiatric clinics.
/rere=uisites or Eecti!eness o
Inter!ie8in) Method:
o Preparation
o /ntroduction
o De.elopin' rapport
o Carr(in' the inter.ie! -or!ard
o Recordin' the inter.ie!
o Closin' the inter.ie!.
?.? /ane( Method:
Pa'e B? o- #10
Data collected -rom the same sample
respondents at inter.al either *( mail or *(
personal inter.ie!.
Hse-ul -or lon'itudinal studies.

Successi.e collection o- data on the same


items.

&eed !ell trained in.esti'ators.


?.+ Mai( Sur!e&:
To collect primar( data *( sendin'
2uestionnaires to respondent
Hse multiple8choice 2uestions, since it is
sel-8administered *( respondents.
?.. E04eri"entation:
To stud( the causal relationship *et!een
dependent and independent .aria*les.
Pa'e BB o- #10
,-ter8onl( research desi'n or *e-ore8a-ter
research desi'n.
E4: Sales promotion, usin' di--erent
ad.ertisin' media.
T(pes o- E4periments:
o Ga*orator( E4periments: E4periment
conducted in a settin' created speci-icall(
-or the purpose
o >ield E4periment: Conducted in real li-e
situation.

Gess o- control o- the researcher.


?.5 Si"u(ation:
, -orm o- o*ser.ational method
The e4periment on the model instead o- the
real s(stem
Simulation: Computer simulation or man
simulation
Pa'e BD o- #10
Suited to stud( the s(stem !ith a set o-
interrelated and independent su*8
components.
?.7 /ro6ecti!e Techni=ue:
Direct method o- data collection .i6 personal
inter.ie!, telephonic inter.ie! and mail
sur.e(: Rel( on respondents o!n *eha.iour,
*elie-s, attitudes etc.
But stimuli to respondents: Picture, a
photo'raph and incomplete sentence.

9pportunit( to capture the real response o-


the respondent.
Three t&4es o /ro6ecti!e Techni=ues:
o )isual: Picture etc
o )er*al: =ord association test, sentence
completion test
o E4pressi.e Techni2ue: i.e some material
and as% to -orm a product.
Pa'e D$ o- #10
E4: Dra!in', role pla(in' etc.
?. Choice o Methods or Data Co((ection:
Case Stud( Method
Samplin' Method.
+. Too(s #Instru"ents$ or Data Co((ection:
Recall di--erent methods o- data collection
=hat tools to collect data:
T(pes o- Tools:
o 9*ser.ation schedule or o*ser.ationnaire
o /nter.ie! schedule
o Document schedule7data sheet
o Nuestionnaire
o Ratin' scale.
Pa'e D1 o- #10

9*ser.ation schedule in o*ser.ation


method, inter.ie! schedule -or inter.ie!in',
2uestionnaire -or mail sur.e( and so on.
&ote: Possi*le o.erlappin' in the use o-
di--erent tools.
..1 Ro(e o Too(s B Instru"ents:
Research o*Fecti.es to ans!er speci-ic
research 2uestions. Response to 2uestions:
Re2uired data to meet research o*Fecti.es.
Data7in-ormation collected: To descri*e the
characteristics o- stud( samples @consumers,
companies, emplo(ees etcA.
=h( need data7in-ormation on
characteristics:: To stud( di--erence in
per-ormance and *eha.iour o- stud( o*Fects.
Data7in-ormation: &eeded to anal(se the
interrelations amon' .aria*les, -or testin'
h(potheses and -or dra!in'
conclusions7recommendations -or decision8
ma%in'.
Pa'e D# o- #10
..2 O1ser!ation Schedu(e #O1ser!ationnaire$:
To record o*ser.ations o- an o*Fect or a
phenomenon
=hat to o*ser.e: &ature and o*Fecti.e o- the
stud( are determinants
Se2uence the items to *e o*ser.ed
Ma%e pro.ision -or .eri-ia*le and
2uanti-ia*le data
Simple units o- o*ser.ation.
..? Ratin) Sca(e:
To measure indi.iduals attitudes, aspirations
and other *eha.ioural aspects.
Measurement Scales @GaterA.
Pa'e D+ o- #10
..+ O4inionnaire:
Gist o- statements pertainin' to an issue or
pro'rame or product7ser.ice
Hse-ul -or conductin' an opinion sur.e(
@opinion pollsA
Respondents to respond to listed
2uestions7statements.
... Docu"ent Schedu(e B Data Sheet:
To o*tain list o- data items -rom documents,
records etc
E4: , stud( o- annual returns and -inancial
statements -iled *( Foint8stoc% companies.
: /mportant /tems: =hen started,
mem*ership, capital structure, net !orth,
.olume o- *usiness, income and
e4penditure, 'ross pro-it, net pro-it *e-ore
ta4 and net pro-it a-ter ta4.
Pa'e D; o- #10
To compare the per-ormance o- di--erent
companies, ha.e uni-ormit( o- items on
!hich data collected.
..5 Schedu(es and @uestionnaires:
Schedule used as a tool -or inter.ie!in'
Nuestionnaire: Hsed -or mailin'
Data collection throu'h Nuestionnaire:
o Cheap and economical

&o -ield sta--


re2uired
o But hi'h rate o- non8response
o /dentit( o- the person responded: &ot
%no!n
o )er( slo!, mailed *ac% dependin' on the
con.enience o- the person
o Can *e used !hen respondents: Giterate
and cooperati.e
Pa'e D< o- #10
o Ris% o- collectin' incomplete and !ron'
in-ormation: "i'h.
Data co((ection throu)h Schedu(e:
o >illed *( the research !or%er or
enumerator

clari-ications on 2uestions
possi*le
o More e4pensi.e
o Go! non8response
o The respondent: Kno!n
o Direct contact !ith the respondent
o /n-ormation7data collected: More accurate.
..7 The ste4s in the Construction o Schedu(e
and @uestionnaire:
The steps almost the same -or *oth.
Determine data needs, dependin' on the
research pro*lem.
Pa'e D0 o- #10
De.elop 1dumm( ta*les3.
Kno!led'e a*out the respondents le.el:
Ge.el o- %no!led'e

choice o- !ords etc.


>or a 2uestionnaire, *etter to ha.e a
structured 2uestions @closed: Res or &oA.
Se2uencin' o- the instrument:
o /denti-ication data
o Brant a!areness
o Brand choice
o Purchase decision
o Brand lo(alt(
o Post8purchase e4perience !ith the
compan(
o Personal in-ormation a*out the
respondent.
E!a(uate the Drat Instru"ent:
Pa'e D? o- #10
o Rele.ance o- the 2uestion to research
o*Fecti.es
o T(pe o- 2uestion: 9pen ended @in.itin'
-ree responseA or closed @Res or &oA
o Clari-(
o &ot %no!led'e demandin'
Pre 5 testin'
=rite do!n instructions7'uidelines -or
enumerator
Res4onse *or" or t&4es o @uestions:
o 9pen8ended Nuestions: Di--erent
responses

Di--icult to classi-(
E4: State reasons -or non8acceptance o-
some products *( consumers.
o T!o8choice or Dichotomous Nuestions:
Res or &o.
Pa'e DB o- #10
o Multiple Choice:
E4: =h( (ou ha.e pre-erred this *rand o-
car:
o Declarati.e Nuestions:
=hich o- the -ollo!in' statements *est
descri*es (our compan(Cs philosoph(:
E4: Ma%e pro-it and sol.e social pro*lems
!hich it ma( create
: Ma%e pro-it, *ut a*ide *( le'al
re2uirements
: ,nd so on.
T&4es o @uestions to 1e a!oided:
o Geadin' 2uestions: Do (ou thin% that the
o.ernment should ha.e more control
o.er corporate sector.
o Goaded Nuestions:
E4: "a.e (ou e.er tried to 'et special
-a.ours -rom a compan( *( pressuri6in':
o ,m*i'uous 2uestions
o Dou*le8*arreled 2uestions:
Pa'e DD o- #10
E4: Do (ou -a.our or oppose producti.it(8
lin%ed !a'e s(stem:
o Gon' 2uestions
o ,.oid dou*le ne'ati.e.
..9 Su""ar& o Characteristics o a ;ood
Schedu(e B @uestionnaire:
9*tain all in-ormation to address research
o*Fecti.es
9nl( rele.ant 2uestions
DonCt as% -or data7in-ormation, !hich can *e
o*tained e--ecti.el( *( methods li%e
o*ser.ation
Do not ha.e unclear, am*i'uous, leadin'
and loaded 2uestions
/- possi*le, a.oid open8ended 2uestions
Nuestions to elicit in-ormation 2uic%l(
Pa'e 1$$ o- #10
Choices to closed 2uestions: ,de2uate and
reasona*le
,rran'e 2uestions in a lo'ical and
ps(cholo'ical se2uence
Code 2uestions and alternati.e ans!er
choices
Pre8test the instrument and de.elop dumm(
ta*les.
..: So"e other Methods o Data Co((ection:
%arrant& Cards:
To o*tain in-ormation re'ardin' sa(
consumer dura*le products.
Distri1utor or Store Audits:
Distri*utors7manu-actures to 'et in-ormation
throu'h their sales persons to estimate
mar%et si6e, mar%et share, seasonal
purchasin' pattern and so on.
Pa'e 1$1 o- #10
/antr& Audits:
/n.esti'ators to estimate consumption o- the
*as%et o- 'oods at the consumer le.el.
Consu"er /ane(s:
Re2uest a set o- consumers to maintain
detailed dail( records o- their consumption.
Mechanica( De!ices:
Hse e(e camera, motion picture camera and
so on in de.eloped !orld to collect re2uired
in-ormation *( *i' companies.
/ro6ecti!e Techni=ues:
,lread( discussed.
De4th Inter!ie8:
To disco.er underl(in' moti.es and desires:
Hsed in moti.ational research.
Content Ana(&sis:
Pa'e 1$# o- #10
o ,nal(sis o- the contents o- documentar(
materials such as *oo%s, ma'a6ines,
ne!spapers.
o Hsed *( researchers at the time o-
literature sur.e(.
Case Stud& Method:
o , -orm o- mainl( 2ualitati.e anal(sis: &ot
merel( 2uantitati.e in-ormation
o Care-ul o*ser.ation o- the stud( unit
o ,n indepth stud( o- processes and
interrelationships.

1The case stud( method is a techni2ue *(


!hich indi.idual -actor !hether it *e an
institution or Fust an episode in the li-e o- an
indi.idual or a 'roup is anal(sed in its
relationship to an( other in the 'roup3.
".
9duman
Pa'e 1$+ o- #10
/hases o Case Stud&:
o Determine the status o- the stud( unit
o Collect data
o Dia'nosis and identi-ication o- casual
-actors o- the pro*lem under stud(
o Su''est remedial measures.
Ad!anta)es:
o , detailed stud(

can o*tain the needed


data7in-ormation in detail
o Possi*le to trace out the historical
*ac%'round o- the stud( pro*lem
o 9- help in -ormulatin' h(potheses
o Can use di--erent methods: Depth
inter.ie!, documents and so on
Pa'e 1$; o- #10
o Enhances the e4perience o- the researcher.
Li"itations o Case Stud& Method:
o /n-ormation collected and
in-erences7recommendations arri.ed at not
compara*le 2uite o-ten !ith other stud(
units
o Time consumin'
o Su*Fecti.it( o- the researcher: Possi*le.
Pa'e 1$< o- #10
MODULE 5
THE SCALIN; TECHNI@UES
1.3 Sca(es in B.R and Cinds o Measure"ent
Sca(es
To measure *eha.iour, %no!led'e and
attitudes.
Cinds o M.S:
No"ina( Sca(e:
o ,ssi'nment o- num*ers to particular
e.ents @Trac% o- consumers, o*Fects etcA
o &um*ers @/.D &oA as la*els -or
identi-ication
o =idel( used in sur.e(s -or classi-(in'
data *( maFor su*'roups o- population.
E4: Hse o- &.S in BR:
Codin': Males J1 >emales J #
Codin' reli'ious a--iliation
Pa'e 1$0 o- #10
: /n horse racin': &.S, sa( num*er ?, to
help *ettors to identi-( that horse.
Ordina( Sca(e:
o To ran% items -rom hi'hest to lo!est
o &o a*solute .alues onl( Ran%in'

Brand , is *etter than Brand B


o Ma'nitude o- di--erence not %no!n
o /n B.R: Respondents to rate career
opportunities, *rands, companies or the
li%e as Qe4cellentC, Q'oodC, Q-airC or
QpoorC

E4cellent hi'her than 'ood, *ut do not


%no! *( ho! much.
Inter!a( Sca(e:
o To indicate order @distanceA and measure
the order @or distanceA in units o- e2ual
inter.als
Pa'e 1$? o- #10
E4: "orse Race:
/nter.al: "
1
"
#
"
+
1 Gen'th # Gen'ths.

9rder o- -inish is %no!n, and the


distance *et!een the horses %no!n.
o Hsed !idel( in *eha.ioural research.
Ratio Sca(e:
o Hsed -or measurin' 2uantitati.e random
.aria*les
o =idel( used in -inancial research, usin'
Rupee7Dollar .alues
o The scale o- measurement is a ratio
o The ratio scales ha.e a*solute 2uantities.
Descripti.e Statistics -or T(pes o- Scales:
T&4e o
Sca(e
Nu"erica(
O4eration
Descri4ti!e
Statistics
Pa'e 1$B o- #10
&ominal Countin'
>re2uenc( in each
cate'or(,
percenta'e in each
cate'or(, mode
9rdinal
Ran%
orderin'
Median, ran'e,
percentile ran%in'
/nter.al
,rithmetic
operations
on inter.als
*et!een
num*ers
Mean, standard
de.iation, .ariance
Ratio
,rithmetic
operation, on
actual
2uantities
eometric mean,
coe--icient o-
.ariation
Co"4osite Measures:
o More than one 2uestion to the respondent
to elicit his opinion on attri*utes o- a
sin'le concept.
o Hse multi8item instrument to measure a
sin'le concept !ith se.eral attri*utes,
called inde4 measures or composite
measure.
Pa'e 1$D o- #10
E41: /nde47composite measure o- a social
class: Based on three !ei'hted
.aria*les: Residence, occupation, and
education.
: Measures o- co'niti.e phenomenon:
Composite /nde4es
E Com*ine them into a composite
measures
E Measurement o- sales personCs
morale: Com*ine 2uestions: "o!
satis-ied are (ou !ith (our Fo*:
"o! satis-ied are (ou !ith (our
territor(: "o! satis-ied are (ou in
(our personal li-e:

,s% di--erent 2uestions to measure


somethin', i.e. morale in this case.
2.3 Criteria *or ;ood Measure"ent: *our
Tests o Sound Measure"ent
Pa'e 11$ o- #10
Re(ia1i(it&: De)ree to 8hich "easures are
ree ro" error and thereore &ie(d
consistent resu(ts.
Similar results o*tained o.er time and
across situations
Repeata*ilit( o- the measure.
E4: Measure Fo* satis-action in t!o period
o- time, sa( 0;M o- the stud( population
satis-ied !ith the Fo*, durin' *oth the
periods o- time.
T8o as4ects o re(ia1i(it&:
o Sta*ilit(: Consistent result !ith repeated
measurements o- the same phenomenon
and !ith the same instrument.
o E2ui.alence ,spect: "o! much error
ma( 'et introduced *( di--erent
in.esti'ators or di--erent samples.
Ho8 to i"4ro!e re(ia1i(it&':
Pa'e 111 o- #10
o Minimi6e e4ternal sources o- .ariation
such as *oredom, -ati'ue

/mpro.ement
in sta*ilit( aspect.
o Hse trained7moti.ated in.esti'ators and
*roaden the sample

impro.ement in
e2ui.alence aspect.
Test o >a(idit&:
,*ilit( o- a measure @-or e4: an
attitude measureA to measure !hat it is
supposed to measure.
E4: Rate o- a*senteeism to measure
morale.
/s rate o- a*senteeism .alid measure o-
morale:
,*senteeism: Due to either a !a.e o-
illiness in the -amil( or unsatis-ied
!or%ers not comin' to !or%:
: Brand Go(alt(: Purchase intensions
1/ !ish / could a--ord Brand L thou'h /
*u( Brand R3.
Pa'e 11# o- #10
Sensiti!it& Test:
Sensiti.it( o- a scale: ,n important
measurement concept.
,n instrumentCs a*ilit( to accuratel(
measure .aria*ilit( in stimuli or responses.
E4: , dichotomous response cate'or(:
,'ree or Disa'ree. ,llo! the recordin' o-
su*tle attitude chan'es. "a.e numerous
items on the scale: Stron'l( a'ree, mildl(
a'ree, neither a'ree nor disa'ree, mildl(
disa'ree, stron'l( disa'ree.

,ll these cate'ories increase a scaleCs


sensiti.it(

/ncrease the sensiti.it( o- a scale *(


allo!in' -or a 'reater ran'e o- possi*le
scores.
Test o /ractica(it&:
Pa'e 11+ o- #10
Jud'e a measurin' instrument in terms
o- econom(, con.enience and
interpreta*ilit(.
Economical: More items 'i.e 'reater
relia*ilit(, *ut more time re2uired -or
inter.ie!.
Con.enience: Measurin' instrument,
eas( to administer.
/nterpreta*ilit(: =hen persons other
than the scale desi'ner are to interpret
results.
?.3 Attitudes and Sca(in) Techni=ues or
Measurin) Attitudes:
?.1 Attitudes:
,n ,ttitude: ,n endurin' disposition to
respond consistenc( in a 'i.en manner to
.arious aspects such as persons, e.ents and
o*Fects.
Pa'e 11; o- #10
Three components o- attitude
o ,--ecti.e Component
E4: 1/ lo.e m( Fo*3 1/ hate oran'e Fuice3

Emotional characteristics o- attitudes.


o Co'niti.e Component: 9nes a!areness o-
and %no!led'e a*out an o*Fect.
E4: Ra.i -eels happ( a*out his Fo*:
1Belie.es that the pa( is 'reat3, or
Ra.iCs Fo* is a 'reat challen'e.
o Beha!ioura( Co"4onent: DBu&in)
intensions and 1eha!ioura( e04ectations
re(ect a 4redis4osition to actionE.

Predisposition to action
E4: Ra.i e4pects to sta( !ith the compan(
and !or% hard to achie.e rapid promotions
-or 'reater .isi*ilit( and in-luence.
?.2 The Re(ationshi4 1et8een Attitudes and
Beha!iour:
o &ot strai'ht -or!ard, thou'h closel(
lin%ed.
Pa'e 11< o- #10
o ,ctual *eha.iour: Ma( *e di--erent -rom
attitudes and *eha.ioural intentions.
?.? Techni=ues or Measurin) Attitudes:
Scalin': ,ssi'nin' num*ers to .arious
de'rees o- opinion, attitude and other
concept.
Scalin': Procedures -or attemptin' to
determine 2uantitati.e measures o-
su*Fecti.e a*stract concepts @E4: ,ttitudesA.
Sca(in) Methods #Techni=ues$:
>i.e 'eneral t(pes o- scales -or
measurin' attitudes
o Ratin) Sca(e:
Re2uires the respondent to estimate
the ma'nitude o- a characteristic or
2ualit( that an o*Fect possesses.
Pa'e 110 o- #10
Respondent to indicate the position on
the scale, !here the( !ould rate the
o*Fect.
E4: ,s% the respondent e.aluate the
st(lin' o- a ne! automo*ile on a -i.e
point scale.
"o! do (ou li%e the product
@Please chec%A
Gi%e Gi%e &eutral
Disli%e Disli%e
.er( some some
.er(
much !hat !hat
much
Can 'o -or 'roup ratin' scale.
o Ran-in) Sca(e
Respondents to ma%e comparisons and
to determine order amon' t!o or more
properties
Choose one alternati.e o.er another
Pa'e 11? o- #10
Ran% order the importance o- com-ort,
er'onomics, per-ormance, and price
-or the tar'et .ehicle
To secure a simple ran%in' o- all items
considered simpl( total ran% .alues
recei.ed *( each item.
o Sortin) Techni=ue
Present respondents !ith se.eral
product concepts, printed on cards
@photosA or .er*al statement o-
product -eatures, such as descriptors
o- the carCs per-ormance.
Re2uire the respondents to arran'e the
cards into a num*er o- -iles

Classi-( the product concepts.


o Cate)oriFation
Put participants @sa( in a car sho!A in
'roups or cate'ories
Pa'e 11B o- #10
,s% them to identi-( their 'ender or
ethnic *ac%'round
/ndicate !hether a particular protot(pe
desi'n !ould appeal to a (outh-ul or
mature dri.er.
o Choice Techni=ue:
,nother t(pe o- attitude measurement
Respondent chooses one o*Fect o.er
another

The respondent pre-ers the


chosen o*Fect o.er the other.
+.3 Attitude Sca(es: T&4es and *eatures
Ratin' Scales: To Fud'e properties o- o*Fects
!ithout re-erence to other similar o*Fects
E4: Do (ou li%e the colour o- a so-tdrin%
Ratin' >orms:
1Gi%e 8 8 8 8 8 8 8 Disli%e
,ppro.e 8 8 8 8 8 8 8 /ndi--erent 8 8 8 8 8 8 8 8
Disappro.e
Pa'e 11D o- #10

Can 'o -or other classi-ications, usin'


e.en more cate'ories.
+.1 Si"4(e Attitude Ratin) Sca(e:
E4hi*it 1#.+. PP ++B8++D.
+.1.1 Si"4(e Cate)or& or Dichoto"ous
Sca(e:
Res and &o
/mportant and Hnimportant
,'ree and Disa'ree.
o Hse-ul -or demo'raphic 2uestions and the
li%es, pro.ide nominal data.
E4hi*it 1#.+. P. ++B
+.1.2 Mu(ti4(e ChoiceG Sin)(e Res4onse
Sca(e:
o Pro.ide nominal data.
Pa'e 1#$ o- #10
+.1.? Mu(ti4(eAChoice and Mu(ti4(eA
Res4onse Sca(e #chec- (ist$:
o Pro.ide nominal data
o Select one or se.eral alternati.es.
+.1.+ Su""ated Ratin)s Method: The
Li-ert Sca(es
o Select a lar'e num*er o- statements.
o Participant to a'ree or disa'ree !ith each
statement.
o Each response is 'i.en a numerical score to
re-lect its de'ree o- attitudinal
-a.oura*leness.
/n e4hi*it 1#.+. P.++B. ,lso see e4hi*it 1#.;
P.+;#.
Pa'e 1#1 o- #10
E4hi*it 1#.+ P.++B
E , scale .alue o- 1: Stron'l( un-a.oura*le
attitude @stron'l( disa'reeA. The other
intensities: # @Disa'reeA, + @&either a'ree
nor disa'reeA, ; @,'reeA, and < @Stron'l(
a'ree i.e a stron'l( -a.oura*le attitudeA.
o ,dd each persons response on a lar'e
num*er o- statements to secure a total score.

GS also called summated scale.


o ,rra( these total score.
o Select some portion representin' the hi'hest
total scores @top 1$M o- the distri*utionA,
and some portion representin' lo!est total
scores @#<M o- the distri*utionA.
o ,nother e4ample
Gi%ert Scale Statements to measure the
ima'e o- the compan(
No
.
State"e
nt
Stro
n)(&
Disa
)ree
Disa
)ree
Neith
er
A)re
e
Nor
A)
ree
Stro
n)(&
A)re
e
Pa'e 1## o- #10
Disa
)ree
1
The
compan(
ma%es
2ualit(
products

#
/t is a
leader in
technolo
'(

+
/t does
not care
a*out the
'eneral
pu*lic

;
The co.
leaders
in RID
to
impro.e
products

<
The co.
is not a
'ood pa(
master

0 The

Pa'e 1#+ o- #10


products
o- the co.
'o
throu'h
strin'ent
2ualit(
tests
?
The co.
has not
done
an(thin'
to cur*
pollution

B
/t does
not care
a*out the
commun
it( near
its plant

D
The cos.
Stoc%s
are 'ood
to *u( or
o!n

1$ The co.
does not
ha.e

Pa'e 1#; o- #10


'ood
la*our
relations
&ote: 1. ,ssi'n numerical .alues to
responses: SD, J 1, D, J #,
&,&D, J +, , J ;, S, J <
#. ,dminister the -ormat to all the
respondents @sa( 1$$A
+. Total up the score.
o The t!o e4treme 'roups: Most -a.oura*le
and least -a.oura*le attitudes to!ards the
attri*ute *ein' studied

T!o criterion
'roups.
o >ollo! item anal(sis: To assesses each item
*ased on ho! !ell it discriminates *et!een
those persons !hose total score is hi'h and
those !hose total score is lo!.
o Calculate the mean score -or each scale
amon' the lo! scorers and hi'h scorers.
Pa'e 1#< o- #10
o >or statistical si'ni-icance test the mean
scores o- *oth the 'roups *( computin' t8
.alues. @Statistical si'ni-icant di--erence
*et!een mean scoresA.
E4hi*it 1#.;. P.+;#.
o Re"e"1er the (i"itation:
Di--icult to %no! !hat a sin'le summated
score means
Man( patterns o- response to the .arious
statements can produce the same total
score.
/dentical scores ma( re-lect di--erent
1attitudes3, *ecause respondents ma(
endorse di--erent com*inations o-
statements.
+.1.. Se"antic Dierentia( Sca(es:
9ri'inall( de.eloped *( Charles 9s'ood
and others.
Pa'e 1#0 o- #10
SD Scale: Based on the proposition that an
o*Fect can ha.e se.eral dimensions o-
connotati.e meanin'.
E4: Roarin' -ire in a -ire place
E Romantic
E Burnin' -lamma*le material !ithin a *ric%
%iln.
,d.ertisers, sales people, product desi'ners
and pac%a'e desi'ners: use !ords, shapes,
associations, and ima'es li%e that.
Presentin' an identi-ication o- a compan(,
product, *rand, Fo* and so on *( a series o-
?8point *ipolar ratin' scales:
E4: Bipolar adFecti.es, 1'ood3 and 1*ad3,
and 1modern3 and old -ashioned, or 1clean3
and 1dirt(3 5 at the *e'innin' and end @or
polesA o- the scale.
Modern 88888 : 88888 : 88888 : 88888 : 88888 : 88888 :
88888 : 9ld >ashioned.
Pa'e 1#? o- #10
/nterpretation -rom le-t to ri'ht:
E4tremel( modern, .er( modern, sli'htl(
modern, *oth modern and old -ashioned,
sli'htl( old -ashioned, .er( old -ashioned,
and e4tremel( old -ashioned.
,ssi'n the !ei'ht to each position
on the ratin' scale:
Sa(, ?, 0, <, ;, +, #, 1
or
S+, S#, S1, $, 81, 8#, 8+
Man( applications in B. R studies:
Per-ormance o- sales persons, 'eneral
characteristics o- the compan(7store, prices
char'ed *( the store, store personnel,
ad.ertisement *( the store, comparison o-
*rands, e--ecti.eness o- ad.ertisin' on
attitude chan'eP comparison o- companies
ima'e.
&ormall( a.oid ne'ati.e scores.
,lso see e4hi*it 1#.<, 1#.0, 1#.?, 1#.B I
1#.D. PP ++; 5 ++0.
Pa'e 1#B o- #10
E4hi*its 1#.+ P.++B
+.1.5 Nu"erica( Sca(e:
E2ual inter.als: E4hi*it 1#.+. P.++D.
9-ten <8point scales, sometimes ? point or
1$ point scales.
Participants: =rite a num*er -rom the scale
ne4t to each item.
Simple: Popular -or mana'ers and
researchers.
E4: )aluatin' the per-ormance a ne!
product @sa( carA please tell us ho! satis-ied
(ou are !ith the car.
E4tremel( Satis-ied ? 0 < ; + # 1 E4tremel(
Dissatis-ied
&umerical Scale utili6es *ipolar adFecti.es
in the same manner as the semantic
di--erential.
Pa'e 1#D o- #10
+.1.7 Mu(ti4(e Ratin) List Sca(e:
Similar to the numerical scale.
But: Circled response -rom the respondent
: Scale Ga(out: >acilitates .isuali6ation.

See E4hi*it 1#.+ P.++D o- the results.


+.1.9 Constant , Su" Sca(es:
To the respondent: Rou ha.e 1$$ points to
distri*ute amon' the -ollo!in'
characteristics o- a restaurant. /ndicate the
relati.e importance o- each attri*ute:
TTTTTTT >ood Nualit(
TTTTTTT ,tmosphere
TTTTTTT Ser.ice
TTTTTTT Price
TTTTTTT KKKK.
TTTTTTT KKKK.
1$$ Total
,lso see e4hi*it 1#.+. P.++D
Pa'e 1+$ o- #10
Di--icult to administer to children or
uneducated.
+.1.: Sta4(e Sca(es:
,n alternati.e to the semantic di--erential,
!hen di--icult to -ind the in.esti'ati.e
2uestion.
See E4hi*it 1#.+ P.++D.
Ratin's ran'e: S< to 8<.
Respondent: Select a plus num*er -or the
characteristic that descri*es the attitude
o*Fect
More accurate the description, lar'er the
S.e num*er
Gess accurate the description, the lar'er
the 5.e num*er.
,nother E4:
Pa'e 1+1 o- #10
=rite do!n Super.isorCs &ame
S+
S#
S1
Supporti.e
81
8#
8+
S S: Similar to S D Scale, *ut easier to use,
especiall( o.er telephone.
S S: Does not re2uire *ipolar adFecti.e

Easier to construct.
But research results -rom *oth SS and SD
Scales, lar'el( same.
Gi%e the Gi%ert, SD, and numerical scales,
SS 'enerall( produces inter.al data.
Pa'e 1+# o- #10
+.1.13 ;ra4hic Ratin) Sca(e:
Respondent: Pro.ided !ith a 'raphic
continuum.
Respondent: Mar% response at an( point
alon' a continuousl(
Score: Measure o- len'th @sa( MillimetersA
-rom either point.
Does not ha.e to *e strai'ht lines use picture
response options7an( other t(pe o- 'raphic
continuum. E4: Gadder Scale
Best Possi*le Gi-e
1$
E2ual distance:
$ to 1,
KKKK , D to 1$
$
=orst Possi*le Gi-e
Pa'e 1++ o- #10
9n !hich run' @11 run'sA o- the Gadder do
(ou -eel (our li-e is to8da(:
,lso see E4hi*it 1#.+. P.++D.
raphic Scales: 9-ten used !ith children.
+.1.11 Su""ar& o Ad!anta)es and
Disad!anta)es o Ratin) Sca(es:
Ratin)
Sca(e
Su16ect
#Res4ondent$
Must
Ad!anta)es Disad!anta)es
1.
Cate'or(
Scale
/ndicate
response
cate'or(
>le4i*le eas( to
respond
/tems ma( *e
am*i'uous !ith
-e! cate'ories,
di--icult to ma%e
realistic
distinctions
#. Gi%ert
Scale

E.aluate
statements on a
scale t(picall(
contains <
alternati.es
Easier scale to
construct
"ard to Fud'e
!hat a sin'le
score means
+. S D and Choose points Eas( to Bipolar
Pa'e 1+; o- #10

&umerical
Scales
*et!een
*ipolar
adFecti.es on
rele.ant
dimensions
construct.
&orms e4ist -or
comparison
adFecti.e must
*e -ound. Data
ma( *e ordinal
not inter.al
;.
Constant
Sum Scale
Di.ide a
constant sum
@sa( 1$$A
amon'
response
alternati.es
,ppro4imates
inter.al
measure
Di--icult -or
respondents
!ith lo!
education le.els
<. Staple
Scale
Choose points
on a scale !ith
a sin'le
adFecti.e in the
centre
Easier to
construct than
SD Scale, eas(
to administer
End points are
numerical, not
.er*al la*els
0. raphic
Scale
Choose a point
on a
continuum
)isual impact,
unlimited scale
points
&o standard
ans!ers
?. raphic
Scale !ith
Picture
Response
Choose a
.isual picture
)isual impact "ard to attach
.er*al
e4planation to
response
Pa'e 1+< o- #10

Cate'ories
+.1.12 Other Ratin) Sca(es:
Thurstone E=ua( , A44earin) Inter!a(
Sca(e: Dierentia( Sca(es
,ttitudes .ar( alon' continua
@Foined7connectedA
Based on consensus scale approach
The researcher to 'ather num*er o-
statement relatin' to the stud( topic
Su*mit the statements, panel o-
Fud'es, !hose ratin' -orms the *asis -or
-inal statements
,dminister these statements to
respondent
=or% out the median .alue o- the
statements chec%ed *( respondents
To 2uanti-( their opinions
Pa'e 1+0 o- #10
Rarel( used in BR, *ecause costl(
and time consumin'.
Measure"ent o Beha!ioura( Intensions
and E04ectations
Cate)or& Sca(e to measure the
respondentCs li%elihood o- action or
intension to per-orm some -uture actions:
E4: "o! li%el( is it (ou !ill chan'e Fo*s
in the ne4t 0 months:
/ de-initel( !ill chan'e
/ pro*a*l( !ill chan'e
/ mi'ht chan'e
/ pro*a*l( !ill not chan'e
/ de-initel( !ill not chan'e
&ote: 9*ser.e the !ordin's o- the
statements !ith 1/3.

Beha!ioura( Dierentia(:
Pa'e 1+? o- #10
o To measure the *eha.iour intensions o-
respondents to!ard an o*Fect or
cate'or( o- o*Fects.
o ,s in the case o- SDS, place the o*Fect
@itemA to *e Fud'ed *( the respondent on
the top.
E4: , #< 5 (ear 5 old >emale
Commodit( Bro%er
=ould 88888 : 88888 : 88888 : 88888 : 88888 :
88888 : =ould &ot
,s% the respondent -or ad.ice.
..3 Ran-in) Sca(es:
Respondent to ma%e comparisons *et!een
t!o or more o*Fects and ma%e the choice.
Ran%in': 9rdinal.
Pa'e 1+B o- #10
E4hi*it 1#.1$ Ran%in' Scales P.+;D
Paired Comparison Scale:
>orced Ran%in' Scale:
Comparati.e Scale:
5.3 Sca(o)ra" Ana(&sis or ;utt"an Sca(o)ra"
Ana(&sis:
, .ersion o- cumulati.e scales.
Statements in the scale -orm a cumulati.e
series

statements are related to one


another. , respondent: Endorsin' one item
re-lectin' the e4treme position results in
endorsin' all items that are less e4treme.
See E4hi*it 1#.; P.+<; and read the
e4planation.
7.3 Mu(tidi"ensiona( Sca(in) and Con6oint
Ana(&sis:
Pa'e 1+D o- #10
Hse-ul methodolo'( -or *rin'in' out
su*Fecti.e Fud'ements o- .arious %inds.
Hsed !hen all the .aria*les are to *e
anal(sed simultaneousl(.
,n important techni2ue -or attitudinal
anal(sis.
But re2uires the %no!led'e o- multi.ariate
statistical anal(sis. E4: Multiple Re'ression
,nal(sis, Discriminant ,nal(sis, >actor
,nal(sis, Cluster ,nal(sis, Canonical
Correlation, Multi.ariate ,nal(sis o-
)ariance @M,&9,A.
Possi*le to use !ith the ad.ancement o-
computer technolo'(.
Pa'e 1;$ o- #10
MODULE 7
SAM/LIN; DESI;NS
1. A. Sa"4(in)G Census and /o4u(ation:
Population: The unit o- stud(
E4: Business Schools in /ndia.
Census: Stud( o- all Business Schools


Complete enumeration o- all items in the
population.
Samplin': Select some o- the Business
Schools, dra! conclusions a*out the entire
population.
1. B. Ste4s in Sa"4(e Desi)n:
Stud( uni.erse
Sample unit
Si6e o- sample
Pa'e 1;1 o- #10
Parameters o- interest
Samplin' procedure

Samplin' desi'n to
decide the num*er o- samples and the
method o- selectin'.
2.3 %h& Sa"4(e'
Go!er cost
reater accurac( o- results
Possi*ilit( o- *etter inter.ie!in'
@testin'A
Better super.ision
Better data processin'
reater speed o- data collection
,.aila*ilit( o- population num*er. /- lar'e
population, samplin' is the onl( process.
?.3 %hat is a ;ood Sa"4(e'
Pa'e 1;# o- #10
Accurac&:
,*sence o- *ias -rom the sample
Hnderestimators o--set the
o.erestimators

, sample .alue that is


close to the population .alue
)ariance as a measure o- error in the
estimate. But, increasin' the sample si6e
can reduce s(stematic .ariance as a cause
o- error.
/recision:
Precision o- estimate
&eed a measure o- ho! closel( the
sample represents the population
Precision measured *( the Standard
Error o- estimate, a t(pe o- standard
de.iation measurement. Smaller the
Standard Error o- estimate, the hi'her is
the precision o- the sample.
Pa'e 1;+ o- #10

Smaller S E E

The ideal sample


desi'n.
Sa"4(in) Error #or rando" sa"4(in)
error$ re(ects the in(uence o chance in
dra8in) the sa"4(e "e"1ers.

Samplin' error: =hat is le-t a-ter all


%no!n s(stematic sources o- .ariance ha.e
*een accounted -or trul( representati.e
sample to 'enerali6e.
Bud'etar( constraint.
+.3 Broad Sa"4(e Desi)ns:
NonA4ro1a1i(it& and /ro1a1i(it&
Sa"4(in):
NonA4ro1a1i(it& Sa"4(in):
o ,r*itrar( and su*Fecti.e
o Each mem*er o- the population: Does
not ha.e %no!n e2ual chance o- *ein'
selected
Pa'e 1;; o- #10
o Researcher: Rel( hea.il( on personal
Fud'ement and con.enience
E4: To -ind out *rand lo(alt( o-
consumers, select internet users as
samples

&o randomness in sample selection.


o &o appropriate statistical techni2ues -or
measurin' random samplin' error -rom
a non8pro*a*ilit( sample.
/ro1a1i(it& Sa"4(in):
o Based on the concept o- random
selection
o Each population element: i.en a
chance -or selection
o Pro.ide estimates o- precision
o >indin's: Could *e 'enerali6ed to
population
o >or E4plorator( Research: &on8
pro*a*ilit( samplin' ma( *e alri'ht.
But -or descripti.e and causal studies,
pro*a*ilit( samplin' needed.
Pa'e 1;< o- #10
>or di--erent t(pes o-
pro*a*ilit( and non8pro*a*ilit( samplin'
desi'ns.
+.1 So"e E(a1orations o /ro1a1i(it&
Sa"4(in):
Si"4(e Rando" Sa"4(in):
o Pro.ides non86ero pro*a*ilit( o-
selection o- population elements


E2ual chance o- selection
o 9nl( one sta'e o- selection
E4: Selectin' the ra--le tic%et -rom a
lar'e drum
o Pro*a*ilit( o- selection J Sample Si6e
Population Si6e
E4: Get: Population si6e J +$$
Sample si6e needed J +$
Pro*a*ilit( o- selection J +$ J
$.1$
Pa'e 1;0 o- #10
E4hi*it 1;.# P.;1D
+.1.1
+$$

The pro*a*ilit( o- selection is


1$M.
o =hen population elements lar'e: Hse
ta*les o- random di'its or computer8
'enerated random num*ers.
Co"4(e0 /ro1a1i(it& Sa"4(in):
o Si"4(e Rando" Sa"4(in): &eeds
population list @Samplin' -rameA,
e4pensi.e and time consumin'

&eed alternati.e samplin' desi'ns:


S(stematic Samplin', Strati-ied
Samplin', Cluster Samplin' and Dou*le
Samplin'.
o S&ste"atic Sa"4(in):
E4: Population: #,$$,$$$ @sa( list o-
names o- companiesA
&eeded sample: 1,$$$ companies

=ith s(stematic samplin', select


e.er( #$$
th
name -rom the list.
Pa'e 1;? o- #10
+.1.2
Select e.er( K
th
element @#$$
th
A
The K
th
element: called s%ip inter.al.
Determine s%ip inter.al @KA
K J Population Si6e
Sample Si6e
E4: Customer accounts can *e
sampled *( usin' the last di'it o-
customer account num*er.
uidelines -or adoptin' s(stematic
samplin':
U /denti-( the list, and num*er the
elements in the population
U /denti-( the s%ip inter.al @KA
U /denti-( the random start
U Dra! a sample *( choosin' e.er(
K
th
entr(.
o Stratiied Sa"4(in):
Di.ide the population into su*'roups,
or strata, *ased on some criterion
E4: Rural and ur*an consumers
"i'h and lo! income 'roups
Pa'e 1;B o- #10
E4: , Pharmaceutical Compan(
interested in measurin' ho! o-ten
ph(sicians prescri*e a certain dru'
Criterion: Ph(sicians
2uali-ication7'ender.
Doctors !ith P De'ree and !ith
onl( BachelorCs De'ree
>urther: Male and >emale Doctors.
>or strati-ied samplin', need to decide
ho! to allocate the total sample
amon' strata.
T!o options -or allocation o- the total
sample:
Proportionate Samplin'
Disproportionate Samplin'.
E4: Proportionate and
Disproportionate Samplin' o-
Business Schools.
H in
/o4u(atio
/ro4ortion
a(
Dis4ro4ortion
a(
Pa'e 1;D o- #10
n Sa"4(e Sa"4(e
"i'h
Ran%
#$M #$M <$M
Moderat
e Ran%
#+M #+M +BM
Go!
Ran%
<?M <?M 1#M
To a.oid o.errepresentation -or lo!
ran% Business Schools, ta%e
disproportionate sample.
Select more samples -rom a strata
!ith 'reatest .aria*ilit( in order to
increase sample e--icienc(, that is,
produce smaller random error.
/ncrease the stratum sample si6e
*ased on 'reatest relati.e .aria*ilit(.
The claims in -a.our o- proportional
strati-ied samplin'
U "i'her statistical e--icienc( than a
simple random sample
Pa'e 1<$ o- #10
U Easier to implement
U PSS: , sel-8!ei'hin' sample.
o C(uster Sa"4(in):
Simple R S: Select each population
element indi.iduals
/n C S: Di.ide the population into
'roups

primar( samplin' unit no


lon'er the indi.idual elements in the
population @e.' manu-acturin' -irmsA
*ut a lar'e cluster o- elements @e.'
citiesA.
Area Sa"4(in): Most 4o4u(ar t&4e
o c(uster sa"4(e
E4: , researcher in retailin': Choose
randoml( se.eral 'eo'raphic areas as
the primar( samplin' unit. Then
inter.ie! all, or a sample o-, retail
stores !ithin the 'eo'raphic clusters

T!o sta'e cluster samplin'.


/deall(, select a cluster 5 a
hetro'eneous unit as the population
Pa'e 1<1 o- #10
itsel- 5 a mirror ima'e o- the
population.
Mu(tista)e Area Sa"4(in) Desi)n:
U E4: Stud( the !or%in' e--icienc( o-
Ban%s in /ndia
U Select districts !ithin the state to
represent di--erent areas
U Select the Bloc%7Talu%
U Select the )illa'es
U Select households -or stud(

>or lar'e8scale area samplin', use


multi sta'e area samplin' desi'n


>re2uentl( used in nation !ide
studies.
o Dou1(e Sa"4(in):
Collect some in-ormation *( sample.
Hse this in-ormation as the *asis -or
selectin' a su*8sample -or -urther
stud(.
Pa'e 1<# o- #10
Called dou*le samplin', se2uencial
samplin' or multiphase samplin'.
E4: Hse a telephone sur.e( to -ind out
interest in Foinin' a ne! Business
School. Strati-( interested respondents
*( de'ree o- interest

su*8sample
amon' them -or intensi.e
inter.ie!in'.
See on
comparison o- Pro*a*ilit( Samplin'
Desi'ns.
..3 NonA/ro1a1i(it& Sa"4(in) Desi)n:
Select stud( units, most con.enientl(
a.aila*le

"i'h pro*a*ilit( o- *ias.


Res &PSD is used:
/- not to 'enerali6e to population &PSD,
o%a(

Gimited o*Fecti.e o- stud(


Go! cost and time
Pa'e 1<+ o- #10
E4hi*it 1;.B P.;+<
PSD calls -or more plannin' includin'
time to contact the randoml( selected
sample
Theoricall( PSD: Superior, *ut -or
practical considerations -ollo! &PSD.
Kinds o- &PS Desi'ns:
Con.enience Samplin':
o >reedom o- researchers to select
!home.er the( -ind
E4: T ) ReporterCs inter.ie! o- person8
on8the street.
o ,lri'ht -or e4plorator( studies.
/ur4osi!e Sa"4(in): T8o t&4es
o Iud)e"ent Sa"4(in):
E4: Stud( o- la*our pro*lems.
/nter.ie! o- those la*our !ho ha.e
e4perienced on8the8Fo* discrimination
Pa'e 1<; o- #10
,ppropriate -or an e4plorator( stud(.
o @uota Sa"4(in):
To impro.e representati.eness.
E4: Samplin' o- male and -emale
students -or a stud( on 2ualit( o-
education in B8School, *ased on the
composition
9r
T!o Cate'ories: Pro-essional and
non8pro-essional de'ree *ac%'round
9r
Reli'ion:
9r
>amil( socio8economic class: Hpper,
middle, lo!er.
Sno81a(( Sa"4(in):
o /nter.ie! the selected respondents.
o 9*tain additional sample respondents,
*ased on in-ormation pro.ided *( initial
respondents.
Pa'e 1<< o- #10
E4: Stud( o- mar%et -or sports
e2uipment or stud( o- consumer
satis-action o- some hi'h8tech consumer
'ood li%e lu4ur( cars.
5.3 Internet Sa"4(in): Se(ect the res4ondents
8ho !isit the co"4an&Js 8e1site.
7.3 The A44ro4riate Sa"4(e Desi)n:
Man( alternati.es, *ut use some criteria to
select the desi'n
De'ree o- accurac(
Costs and time needed
,.aila*ilit( o- population elements
&ational .ersus local research proFect
&eed -or statistical anal(sis
E4: &PS Desi'ns do not allo! the
researcher to utili6e statistical anal(sis,
!hereas statistical anal(sis possi*le !ith
PSD.
Pa'e 1<0 o- #10
reater speed o- data collection.
9.3 Sa"4(in) *ra"e:
Closel( related to the population
Gist o- elements -rom !hich sample is
dra!n.
:.3 Sa"4(e SiFe:
See ,ppendi4 1; a PP ;;+8;<#
Some m(ths:
Gar'e sample to *e representati.e
Proportional relationship to the si6e o-
the population

Possi*le in &PSD.
/n PSD: "o! lar'e the sample: , -unction
o- .ariation in population parameters and
precision needed *( the researcher.
Pa'e 1<? o- #10
Keep some *road principles to determine
sample si6e: @see ,ppendi4 1<aA
reater the .ariance !ithin the
population, the lar'er the population to
ensure precision in estimation.
Gar'er sample to ensure the 'reater
precision in estimation.
Gar'er sample to ensure smaller error
ran'e.
Gar'er sample, i- 'reater num*er o-
su*'roups !ithin the sample

Each su*'roup to ha.e minimum


sample si6e.
Cost considerations to decide the data
collection methods and sample si6e.
Pa'e 1<B o- #10
MODULE 9
DATA EDITIN;G CODIN; AND ENTERIN;
1.3 Data /re4aration: EditG Code and enter the
co((ected data

Prepare the ra! data -or anal(sis.


1.1 The Ne0t Ste4: Descripti.e statistical
summar( to understand the collected data.
Details in Module D.
2.3 K Editin):
To 'uarantee that data are
o ,ccurate
o Consistent !ith the intent o- 2uestion
o Hni-orml( entered
o Complete
o ,rran'ed to simpl( codin' and
ta*ulation.
>ield Editin': B( -ield super.isors
Pa'e 1<D o- #10
Central Editin' @/n8"ouse Editin'A
uidelines to Data Editors:
o >amiliar !ith instructions to
inter.ie!ers7coders
o Maintain ori'inal entr( in tact
o Hse distincti.e colour -or ma%in'
editin' entries
o /nitial and put data editin' data
o Edit -or consistenc(:
E4: Decision8collect data onl( -rom 1B
(ears o- old and a*o.e. /- -ound
1?710K. Rear old respondent, chec% out
that sample unit.
o Edit -or completeness:
E4: Does (our or'anisation ha.e more
than one internet !e* site:
Res TTTTTTTTTT &o TTTTTTTTTTT
/- Res, ho! man( TTTTTTTTT
/- mentioned, + !e*sites !ithout
chec%in' either Res or &o: =hat to
do:
Pa'e 10$ o- #10
o Edit -or -acilitatin' codin'.
?.3 Codin):
C: ,ssi'n num*ers7other s(m*ols to
responses7ans!ers.
E Possi*le to partition cate'ories @*ased on a
'i.en .aria*leA
E Sa(, $ -or male, 1 -or -emale
E =ill *e o- help in the use o- dumm(
.aria*le techni2ue to capture the net e--ect
o- 'ender. @Re-erence in P.0?#A.
E Hse o- d.) method@E4.A.
Codes: Data can *e processed in a computer.
?.1 K Codes: Or)anise data into ie(dsG records
and i(es.
>ield: , collection o- characteristics
5 , sin'le num*er, letter o- the alpha*et
etc.

Sin'le t(pe o- data.


Pa'e 101 o- #10
Record: Collection o- related -ields.
>ile: , collection o- related records.
?.2 K Data Matri0: Rectan'ular arran'ement o-
data in ro!s and columns
E4: Spread Sheet.
?.? K Code1oo- Construction #Codin)
Sche"e$:
Contains each .aria*le in the stud(
and speci-ies the application o- codin'
rules to the .aria*le
>acilitates more accurate and
e--icient data entr(
Code*oo%s: Nuestion num*er,
.aria*le name, location o- the .aria*les
code on the input medium.
E4: Spread Sheet or SPSS data -ile.
E4hi*it 1<.#. P.;0#.
Pa'e 10# o- #10
?.+ K To Basic Ru(es or Code Construction:
E4hausti.e codin' cate'ories: ,ll
su*Fects or o*Fects or responses
Mutuall( e4clusi.e and independent

&o o.erlap *et!een cate'ories.


Prerecordin' @/n the NuestionnaireA and
Post8recordin' @a-ter data collectionA.
?.. K Codin) C(osed @uestions:
,nother E4ample:
1.a Do (ou personall( o!n corporate
stoc%s
1 #
1.* Do (ou o!n stoc%s in the corporation
-or !hich (ou !or%:
Do (ou o!n then in a corporation -or
!hich (ou do not !or%.
Pa'e 10+ o- #10
E4hi*it 1<.+ P.;0+
Res &o
1 Compan( -or !hich / !or%
# 9ther compan(.
@Please as man( as appl(A.
?.5 Codin) O4en , Ended #*ree Res4onse$
@uestions:
Researcher: &o clear h(potheses re'ardin'
the ans!ers
,ns!ers @ResponsesA: &umerous and .aried
Codin' in this case: Reduce the lar'e
num*er o- indi.idual responses to a -e!
'eneral cate'ories that can *e assi'ned
numerical codes -or codin' closed 2uestions
@1, #, +, ; and <A.
?.5.1 Content Ana(&sis *or O4enAEnded
@uestions:
E4: Me'a Trends: Based on articles7ne!s in
ne!s papers -or a decade or so.
Pa'e 10; o- #10
+.3 Data Entr&: >or details see P ;?; 5 ;?B.
Con.ert data7in-ormation to a medium -or
.ie!in' and manipulation. Three maFor
-ormats -or data entr(:
Ke(*oardin'
Bar Codin'
9ptical Character Reco'nition @9CRA
Man( on the hori6on.
Pa'e 10< o- #10
MODULE :
STATISTICAL DESCRI/TION O* DATA:
TABULAR
..3 K Recall data editin', codin' and enterin'
sta'es o- research @Module BA.
Create numerical summaries o- entered data

Descripti.e statistical tools -or cleanin'


data, disco.erin' pro*lems, and
summari6in' distri*utions.
Data Distri*ution: ,n arra( o- .alue counts
-rom lo!est to hi'hest .alue o- a .aria*le,
*ased on ta*ulation.
5.3 K Descripti.e Statistical Measures: To depict
the centre, spread, and shape o-
Distri*utions.
E Preliminar( tools -or data description,
called descripti.e statistics.
E Descripti.e Statistics: "elp-ul in
trans-ormin' ra! data into a -orm so as to
Pa'e 100 o- #10
-acilitate understandin' and interpretin',
rearran'in', orderin' and manipulatin'
data to pro.ide descripti.e in-ormation.
7.3 K Di--erent Kinds o- Measurements and
T(pes o- Descripti.e Statistics:
Ta*ulation: >re2uenc( distri*ution,
percenta'e and cumulati.e percenta'e.
Measures o- Central Tendenc(:
o Mean
o Median
o Mode.
Measures o- )aria*ilit(:
o )ariance
o Standard De.iation.
Measures o- Shape:
o S%e!ness
o Kurtosis.
Pa'e 10? o- #10
7.1 K Ta1u(ation:
E4 1 : GCD T) Mar%et
: Sur.e( o- D manu-acturers @D *randsA
to -ind out M unit sales increase.
:
E >re2uenc( distri*ution percenta'e
and cumulati.e percenta'e.
E >re2uenc( distri*ution *( -orei'n
and domestic companies.
7.1.1 K Standard Nor"a( Distri1ution:
A S4ecia( Case o ND
Manu-acturer: E4perienced ?M
increase in annual unit sales
?J mean o- the GCD T) distri*ution
Pa'e 10B o- #10
E4hi*it 10.# P.;D+
Standard score @or V scoreA to
indicate ho! man( manu-acturin' units:
,*o.e or *elo! the mean
V score is the Standarised score:
E4: Manu-acturin' unit 6enith has
e4perienced an increase in annual unit
sales o- DM
)alue8Mean
VenithCs Standard Score @or V scoreA J
888888888888888888888888888888
Standard
De.iation
J D 5 ? J 1.0;
1.##
7.1.2 K Cross , Ta1u(ation:
To ans!er man( research 2uestions
E4: Nuestion: Do (ou shop at R 5
Mall:
Cross Ta*ulations:
Les No Tota(
Pa'e 10D o- #10
Men 1<$ ?< ##<
=omen 1B$ ;< ##<
Tota( ??3 12. +.3
Percenta'e Cross8Ta*ulation o- Nuestion
@Do (ou shop at R8MallA. B( Se4 o-
Respondents, Ro! M
Les No Tota(
Men 00.?M ++.+M
1$$M
@:A
=omen B$.$M #$.$M
1$$$M
@:A
Percenta'e Cross8Ta*ulation o- Nuestion
@Do (ou shop at R8MallA, Column M
Les No
Men ;<.<M 0#.<M
=omen <;.<M +?.<M
Tota( 133.3 133.3
Cross8Ta*ulation o- Marital Status, Se4,
and Response to the Nuestion: Do (ou
shop at R8Mahal
Do Lou Married Sin)(e
Pa'e 1?$ o- #10
Sho4
at LA
Maha(
Men %o"en Men %o"en
Res <<M B$M B0M B$M
&o ;<M #$M 1;M #$M
7.1.? K @uadrant Ana(&sis: >ariation o
the CrossATa1u(ation:
Popular as a component o- total
2ualit( mana'ement @TNMA pro'rammes.
>our 2uadrants on a t!o8dimension
ta*le.
/n B.R: Plots the relationship
*et!een the a*out a product attri*uteCs
importance and a.era'e ratin's o- a
compan(Cs @*randCsA per-ormance !ith
respect to that product attri*utes.
/mportance8Per-ormance ,nal(sis:
Rate ho! important a particular
attri*ute: Rate the per-ormance o- a
particular *rand !ith respect to that
attri*ute.
Pa'e 1?1 o- #10
Nuadrant ,nal(sis -or a Micro!a.e
Meal:
"i'h /mportance7 "i'h
/mportance7
Go! Ratin' "i'h Ratin'
?
0
<
;
+
#
1
1 # + ; < 0 ?
Go! /mportance7
Go! /mportance7
Go! Ratin' "i'h Ratin'
Ratin)s o /roduct /eror"ance on
Each Attri1ute
Pa'e 1?# o- #10
?, 1 @Eas( to prepareA
0,+ @&utritiousA 0,0 @ood tasteA
<,# @>amil( MealA <,< @>illin'A
#,1 @>anc( MealA #,< @Gate &i'ht Snac%A

o Hpper8le-t and upper8ri'ht 2uadrants


and lo!er8le-t and lo!er8ri'ht
2uadrants. =hat do the( indicate:
/mportance8Per-ormance Ratin':
o E4 : The Hpper8Ri'ht Nuadrant:
/mportance ratin' a*o.e ; @i.e 0 and
<A, and per-ormance ratin' a*o.e ;
@i.e 0 and <A.
: Similarl( interpretation o- other
2uadrants.
Data Trans-ormation:
Data con.ersion: Chan'e data -rom
their ori'inal -orm to a -ormat that *etter
supports data anal(sis to achie.e research
o*Fecti.e.
E4: Collapsin' a >i.e8Cate'or( Gi%ert
Scale
Li-ert Sca(e as it a44ears in the @uestionnaire
/ncreased -orei'n in.estment in /ndia poses a
threat to our economic independence
Pa'e 1?+ o- #10
o Stron'l( ,'ree o ,'ree o &either ,'ree
o Disa'ree o Stron'l( Disa'ree
&or Disa'ree
Ta*ulation o- Data in 9ri'inal and Collapsed
)ersions
. , /oint /ercenta)e
Co((a4sed
Sca(e
/ercenta)e
Stron'l(
,'ree
+
Stron'l(
,'ree7
,'ree
1<
,'ree 1#
&either
,'ree
&or
Disa'ree
+$
&either
,'ree
&or
Disa'ree
+$
Disa'ree ;<
Stron'l(
Disa'ree7
Disa'ree
<<
Stron'l(
Disa'ree
1$
/nde4 &um*er: /n B.R -re2uentl(
used -or research !ith secondar( data.
Recall the concept o- CP/ and =P/: >or
Pa'e 1?; o- #10
comparin' price chan'e o.er a period o-
time.
7.1.+ K Ca(cu(atin) Ran- Order
Ta*le: /ndi.idual Ran%in' o- Dream
Destinations
/erson Ha8ai /aris ;reece China
1 1 # ; +
# 1 + ; #
+ # 1 + ;
; # ; + 1
< # 1 + ;
0 + ; 1 #
? # + 1 ;
B 1 ; # +
D ; + # 1
1$ # 1 + ;
To 9*tain Ran% 9rder:
"a!ai: @+L1A S @<L#A S @1L+A S @1L;A J
#$
Paris: J #0
reece: J #0
China: J #B
Pa'e 1?< o- #10
Ta*le: >re2uenc( Ta*le o- Dream
Destination Ran%in's
Destinat
ion
/reerence Ran-in)
1
st
2
nd
?
rd
+
th
"a!ai + < 1 1
Paris + 1 + +
reece # # ; #
China # # # ;
E The lo!est total score

-irst @hi'hestA
pre-erence ran%in' and hi'hest total score

last @lo!estA pre-erence ran%in'.


E Ran% 9rderin': 1. "a!ai
#. Paris
+. reece
;. China.
9.3 ;ra4hic Methods o Dis4(a&in) Data:
Gine raphs: E4: Comparisons o.er time
L8a4is: Time
R8a4is: Price o- Shares o-
di--erent companies.
Pie Dia'ram
Pa'e 1?0 o- #10
Bar Dia'ram
Computer raphics7Computer Mappin'.
:.3 Lou are a(read& a"i(iar 8ith Measures o
Centra( Tendenc&G Measures o >aria1i(it&
and Measures o Sha4e.
Pa'e 1?? o- #10
MODULE 13
HL/OTHESIS TESTIN;
13.3 K H&4othesis: An un4ro!en
4ro4osition or su44osition
: , statement, an assumption a*out
the nature o- the
phenomenon.
E4 : R I D Mana'erCs "(pothesis:
En'ineers hi'hest in technical
%no!led'e !ill *e the most
producti.e.
: , personnel Mana'erCs "(pothesis:
, positi.e chan'e in attitudes to!ards
Fo* securit( !ill lead to an increase in
emplo(ee retension.
13.1 K H&4othesis Testin) and Research
/rocess
Pa'e 1?B o- #10
E4hi*it 1?.1 P.<#B
13.2 The Nu(( H&4othesis and A(ternati!e
H&4othesis:
Researchers: E4tremel( cautions in testin'.
"ence state the statistical h(pothesis in a
null -orm.
E4 1: The case o- the -uel e--icienc( and o- a
car !ith some modi-ications.
&. ": The population parameter o- <$
mp' has not chan'ed !ith modi-ication
"
9
:

J <$ mp'.
,. ": The population parameter o- <$
mp' has chan'ed
"
,
:

<$ mp'.
E4 #: "i'hl( do'matic @ie closed mindedA
indi.iduals: Gess li%el( to tr( an inno.ati.e
mana'ement techni2ue than less do'matic
indi.iduals.
"
9
: There is no di--erence *et!een
1hi'h do'matics and lo! do'matics3A in
their !illin'ness to tr( inno.ations.
Pa'e 1?D o- #10
"
,
: There is a di--erence *et!een hi'h
and lo! do'matics

opposite o- "
9
.
13.? H&4othesisATestin) Statistica( /rocedure:
State the "
9
Choose the appropriate statistical test @ch 1B
o- CIPA
Select the desired le.el o- si'ni-icance
Compute the calculated di--erence .alue
9*tain the critical test .alue
/nterpret the result: /- the calculated .alue W
critical .alue

ReFect "
9
.
/- calculated .alue X critical .alue

>ailed
to reFect the "
9
.
13.+ T&4e I and T&4e II Errors:
Pa'e 1B$ o- #10
T(pe / Error: Decision: ReFect the "
9
, e.en
thou'h true.
, true "
9
reFected.
The researcher concludes
that there is a statisticall(
si'ni-icant di--erence !hen
in realit( one does not e4ist.
T(pe // Error: Decision: "
,
true, *ut not to
reFect "
9
.

>ails to reFect a -alse "


9
: Correct Decision: /- "
9
is -alse,
reFect it.
13.. /ara"etric >ersus Non4ara"etric
H&4othesis Tests:
Di--erence *et!een the t!o tests: Depends
on underl(in' a*out the data to *e anal(sed.
Three t(pes o- Data ,nal(sis:
Hni.ariate Data ,nal(sis: To assess the
statistical si'ni-icance o- a h(pothesis
a*out a sin'le .aria*le
Pa'e 1B1 o- #10
Bi.ariate Data ,nal(sis: The statistical
si'ni-icance test used in the simultaneous
in.esti'ation o- t!o .aria*les, usin' tests
o- di--erences or measurement o-
association *et!een t!o .aria*les at a
time.
Multi.ariate Data ,nal(sis: Statistical
methods that allo! the simultaneous
in.esti'ation o- more than t!o .aria*les.
/ara"etric H&4othesis Tests:
Data: /nter.al or ratio8scale
Sample Si6e: Gar'e
Stud( data dra!n -rom
populations !ith normal @*ell shapedA
distri*ution and7or normal samplin'
distri*utions.
E4: Data on &P and industr(
sales .olume

Comparison o- the inter.als.


Pa'e 1B# o- #10
Possi*le Statistical Tests: V8test
@or t8testA -or a h(pothesis a*out a mean,
product8moment correlation anal(sis,
anal(sis o- .ariance tests Chi8S2uare8test
and >8test.
Non4ara"etric H&4othesis Tests:
&o assumption o- normalit(
Data: Either ordinal or nominal
Data anal(sis o- *oth nominal
and ordinal scales t(picall( uses
nonparametric statistical tests.
E0a"4(es o Se(ectin) the A44ro4riate
Uni!ariate Statistica( Method:
Business
/ro1(e"
Statistica(
@uestion
/ossi1(e Tests
o Statistica(
Si)niicance
/nter.al or Ratio
Scale: Compare
actual and
h(pothetical
.alues o- a.era'e
/s the sample
mean
si'ni-icantl(
di--erent -rom the
h(pothesi6ed
V8test: >or
lar'e
population
T8test: >or
small
Pa'e 1B+ o- #10
salar( population mean population
9rdinal Scale:
Compare actual
and e4pected
e.aluations
Does the
distri*ution o-
scores on a scale
!ith the
cate'ories
e4cellent, 'ood,
-air and poor
di--er -rom the
e4pected
distri*ution
Chi8S2uare test
Determine
ordered
pre-erence -or
*rands in a
product class
Does a set o-
ran% orderin's in
a sample di--er
-rom an e4pected
or h(pothetical
ran% orderin':
Kolmo'oro. 5
smirno. test
&ominal Scale:
/denti-( se4 o-
%e( e4ecuti.es
/ndicate
percenta'e o-
%e( e4ecuti.es
!ho are male
/s num*er o-
-emale
e4ecuti.es
e2ual to the
num*er o- male
e4ecuti.e:
/s the
proportion o-
Chi8S2uare test
t8test o- a
proportion
Pa'e 1B; o- #10
male e4ecuti.es
the same as the
h(pothesi6ed
proportion:
In Brie:
t8test and V8test -or small and lar'e samples
respecti.el(, !ith normall( distri*ute
population7sample
Chi8s2uare test: Test statistical si'ni-icance
in the anal(sis o- -re2uenc( distri*utions
Test o- a proportion : t8test or V8test
Main parametric tests: V8test or t8test
Main non8parametric tests: Chi8s2uare test.
Pa'e 1B< o- #10
MODULE 11
HL/OTHESIS TESTIN;: METHODS O*
STATISTICAL TESTIN;
11.3 K Reca(( three "ethods o data
ana(&sis:
Hni.ariate data anal(sis
Bi.ariate data anal(sis
Multi.ariate data anal(sis
,lso recall the di--erence *et!een
parametric and nonparametric h(pothesis
tests.
Re-: Ch. 1?.
12.3 One , Sa"4(e Test: /ara"etric and
Non4ara"etric Tests.
12.1 K /ara"etric Tests:
V8test or t8test used, !hen
population7sample normall( distri*uted
Pa'e 1B0 o- #10
Thum* Rule: Hse V8test, !hen &

1#$ @Gar'e sampleA. >or small sample, ie


&

+$, use t8test


Some applications o- one8sample
test: P.<$0 o- C I P
E4: The h(*rid .ehicle pro*lem -or
statistical procedure: PP:<$0 and <$? o- C
I P.
12.2 Non4ara"etric Test:
)ariet( o- non8parametric tests, dependin'
on the measurement scale.
>or nominal measurement scale: Hse either
*inomial test or the chi8s2uare one8sample
test.
Binomial Test:
=hen, t!o classes o- population sample:
E4: Male and -emale
: Bu(er and non8*u(er.
Pa'e 1B? o- #10
>or small sample.
Chi8S2uare Test:
=idel( used particularl( in case o-
nominal data, !ith hi'her classes, ie more
than t!o classes o- population7sample.
>or procedure to use chi8s2uare PP 0<$
5 0<+.
1?.3 T8o , Inde4endent , Sa"4(es Tests:
T!o independent samples:
E4 : ,nnual salaries o- pro-essional
-emale emplo(ees and those o-
pro-essional male emplo(ees: Di--erence:
statisticall( di--erent:
: Be'innin' salar( pac%a'e o--ered to
-emale MB, candidates and male MB,
candidates in a 'i.en Business School.
Pa'e 1BB o- #10
: Chan'es in the mar%et share o- t!o
companies, sellin' competin' products.
: The impact o- t!o trainin' methods on
sales.
Bi.ariate Statistical ,nal(sis: Tests o-
di--erences, sa( *et!een t!o means.
1?.1 /ara"etric Tests:
V8test @-or lar'e sampleA and t8test @-or small
sampleA -re2uentl( used. Can also use >8test.
"(pothesis to test the impact o- t!o trainin'
methods @, I BA on sales:

"
9
: L
,
J L
B

"
,
: L
,
W L
B
or "
,
: L
,

L
B
.
1?.2 Non4ara"etric Tests:
Chi8S2uare: To test -or di--erences *et!een
t!o samples.
Pa'e 1BD o- #10
Chi8S2uare: )alua*le -or nominal data, *ut
can also *e used !ith ordinal measurements.
1+.3 T8o , Re(ated , Sa"4(e Tests:
E4 o- Related Samples:
Compare the consumption o- hus*ands
and !i.es
Per-ormance o- emplo(ees *e-ore and
a-ter .acations
Pre8trainin' and post8trainin'
per-ormance o- sales persons
Per-ormance o- students *e-ore and a-ter
the discussion o- a Module in Business
Research.
Both parametric and non8parametric tests
applica*le.
Pa'e 1D$ o- #10
1+.1 /ara"etric Tests:
t8test -or independent samples onl(, *ut here
t!o8related samples.
"o! to reduce t!o sample case to the
e2ui.alent o- one8sample case, and to
-ormulate h(pothesis -or testin', usin' t8test
@called paired8samples t8testA.
1+.2 Non4ara"etric Tests:
Mc&emar Test: =ith either nominal or
ordinal data
M T: Hse-ul !ith *e-ore8a-ter measurement
o- the same su*Fects
1..3 C , Inde4endent , Sa"4(e Tests:
Three or more samples in.ol.ed
To learn !hether the samples ha.e come
-rom the same populations
Pa'e 1D1 o- #10
Hse ,&9), and the >8test, !hen data
measured on an inter.al8ratio scale
Hse nonparametric test, i- data measured on
an ordinal or nominal scale.
1..1 /ara"etric Tests:
E4: Means o- se.eral populations are e2ual:
,&9),
OneA8a& ana(&sis o !ariance:
Compare the e--ects o- one treatment o-
-actor8sa( e--ects o- t(pes o- retail stores
on shoppin' e4penditure.
o The test statistics -or ,&9),: > Ratio
o Details: PP:<108<#$ on ratin' o- in
-li'ht ser.ice ratin's o- three ,irlines *(
tra.elers

>actor: ,irline
Dependent: >li'ht ser.ice ratin'
o "(pothesis:
"
9
:

,1
J

,#
J

,+
: The means @the in
-li'ht ser.ice ratin'sA are e2ual
Pa'e 1D# o- #10
"
,
:

,1


,#


,+
: The means are not
e2ual.
T8o %a& ANO>A:
o To learn more a*out ser.ice ratin's,
select one more -actor: Seat selection o-
passen'ers in addition to the -irst -actor,
i.e airline
o Three 2uestions to *e considered in one
model:
,re di--erences in -li'ht ser.ice
ratin's attri*uta*le to airlines:
,re di--erences in -li'ht ser.ice
ratin's attri*uta*le to seat selection
Do the airline and the seat selection
interact !ith respect to -li'ht ser.ice
ratin's:
1..2 Non4ara"etric Tests: O4tiona(.
Pa'e 1D+ o- #10
15.3 C , Re(ated , Sa"4(e Tests:
17.3 Measures o Association:
Man( mana'ement 2uestions around
relationships *et!een t!o or more .aria*les.
Relationship h(pothesis re2uired to
ans!er these mana'ement 2uestions.
E4 : /s sales producti.it( associated !ith
pa( incenti.es:
: /s socioeconomic status associated !ith
the li%elihood o- purchasin' o-
recreational .ehicle:
: =hat is the relationship *et!een the
candidates ran% *ased on an e4ecuti.e
inter.ie! and the ran% o*tained -rom
testin':
: =hat is the relationship *et!een mar%s
o*tained *( students in internal
assessment and mar%s o*tained in -inal
e4amination:
Pa'e 1D; o- #10
: Some companies ha.e *een a*le to
increase their sales *( #$M *( increasin'
their ad.ertisement e4penditure. Can !e
predict ne4t (earCs sales *ased on present
ad.ertisement e4penditure:
The -irst three 2uestions: To *e ans!ered
*( measures o- association @correlationA.
The last 2uestion: To *e ans!ered !ith
simple linear re'ression.
Correlation: , measure o- nature o-
association *et!een t!o .aria*les !hich is
the cause and !hich is the e--ect, not %no!n

Correlation does not mean causation.


Re'ression: Predict the .alues o- a
dependent .aria*le !ith chan'es in the .alue
o- independent @e4planator(A .aria*les


cause 5 e--ect relationships, %no!n.
17.1 Bi!ariate Si"4(e Corre(ation Ana(&sis:
/ara"etric Corre(ation
Pa'e 1D< o- #10
&eeds t!o continuous .aria*les measured
on an inter.al or ratio scale
Y J Pearson Correlation Coe--icient.
The C.coe--icient: Does not distin'uish
*et!een independent and dependent
.aria*les.
The .aria*les treated s(mmetricall(

Both
Y
4(
and Y
(4
ha.e same interpretation.
,ssociations *et!een t!o .aria*les: Ginear

Strai'ht line association, e4ceptin' Y J $.


Y : Re.eals the ma'nitude and direction o-
relationships
Y : )aries o.er a ran'e o- S 1 throu'h $ to 5
1
Scatter plots -or e4plorin' the
relationships7associations:
/nterpretation o- Y:
Pa'e 1D0 o- #10
&o e.idence o- cause and e--ect, e.en
!hen Y is hi'h and statisticall( si'ni-icant.
Se.eral e4planations -or correlation
results:
o L causes R
o R causes L
o L and R: ,cti.ated *( one or more
other .aria*les
o L and R: /n-luence each other
reciprocall(.
Spurious correlation:
E4: Correlation *et!een teacherCs salar(
and the consumption o- li2uor o.er a
period o- (ears.
Get YJ$.D

Does not indicate that


teachers drin%, nor does it indicate that the
sale o- li2uor increases teacherCs salaries.
Gi2uiror sales and teachers salar( co.ar(
*ecause the( are in-luenced *( a third
Pa'e 1D? o- #10
.aria*le, sa( lon'8run 'ro!th in national
income7population.
&ote: &onparametric measures o-
association: &ot discussed. &eeds the
%no!led'e o- Spearman ran%8order
correlation coe--icient method, and the
Kendall Ran% correlation coe--icient
method.
17.2 Si"4(e Linear Re)ression:
9ne dependent and one independent
.aria*le: The Model
R
i
J Z
$
S Z
1
L
i
S [
i

/nterpretation o- the model.


Estimation procedure and testin' the
'oodness o- -it as in C I P.
Coe--icient o- determination: R
#
, measure o- proportion o- .ariance in
R e4plained *( L
, measure o- linear relationship
Pa'e 1DB o- #10
,n important indicator o- the predicti.e
accurac( o- the e2uation: Thum* Rule: R
#
W B$M or more to *e a relia*le indicator o-
prediction

Gess than B$M

predicti.e
accurac( decreases.
Pa'e 1DD o- #10
MODULE 12
RESEARCH RE/ORT /RESENTATION:
ORAL AND %RITTEN
19.3 The Ma6or Deter"inants o @ua(it& o
Research:
,ddressin' mana'ement dilemma,
mana'ement 2uestions, in.esti'ati.e
2uestions, research desi'n, and research
report.
Poor -inal report or presentation:
Bad 2ualit( o- research e--ort
Some times, technical details ma( *e
o.erloo%ed
E But research report is the crucial
means -or communicatin' the !hole
proFect, impactin' the compan( decision.
"ence the Nualit( o- Research Report:
Crucial.
Pa'e #$$ o- #10
1:.3 R
e4ortin) 4hase o the research 4rocess.
23.3 T&4es o Research Re4orts:
Short Re4orts:
o Pro*lem: =ell de-ined, !ith limited scope
o Simple and strai'ht -or !ard methodolo'(
o S R: /nterim reports to pro.ide
in-ormation7e4tent o- pro'ress
o ,*out < pa'es: , *rie- statement a*out the
authori6ation -or the stud(, the
pro*lem e4ercised and scope o-
the stud(

: Conclusions and
recommendations, and -indin's
that support them.
o , letter o- transmittal
Pa'e #$1 o- #10
E4hi*it #$.1 Pa'e ?$D
o Ma%e ample use o- 'raphics
o >i.e 'uidelines -or short reports: Pa'e
?1$.
T8o T&4es Lon) Re4orts: Technica( and
Mana)e"ent Re4ort
o The Technica( Re4ort: /ncludes -ull
documentation and detail
: Details on !hat !as done,
ho! done.
o /nclude sources o- data, research
procedures, samplin' desi'n, data
'atherin' instruments, inde4
construction, and data anal(sis methods:
Some in ,ppendi4.
o Conclusions and recommendations:
Deri.ed -rom speci-ic -indin's.
o , *rie- re-erence to other research,
theories and techni2ues.
Pa'e #$# o- #10
The Mana)e"ent Re4ort:
o >or the non8technical client
o /n.ert the order o- presentation
o ,-ter the pre-ator( section and
introductor( section:
Conclusions and recommendations
/ndi.idual -indin's ne4t
The appendices: Methodolo'ical
details
Ma%e e4tensi.e use o- .isual displa(s.
21.3 %ritten Research Re4ort Co"4onents:
The ;enera( *or"at or Or)anisin) the
Re4ort.
21.1
Pa'e #$+ o- #10
E4hi*it #$.# P.?1#
E Research Report Sections and Their 9rder
o- /nclusion
E Details in PP:?1# 5 ?1?
E See E4hi*it #$.+ -or reportin' -indin's.
21.2 So"e other Considerations or I"4ro!in)
the @ua(it& o Re4ort:
Reada*ilit(
Comprehensi*ilit(
Tone
E--ecti.e use o- ta*les and 'raphic aids @E4:
Pie ChartsA.
21.? An A(ternati!e ;enera( *or"at or
Re4ort %ritin):
1.Title pa'e
#.Getter o- transmittal
+.Getter o- authori6ation
Pa'e #$; o- #10
;.Ta*le o- contents and lists o- ta*les and
-i'ures
<.E4ecuti.e Summar(:
9*Fecti.es
Results
Conclusions
Recommendations.
0.Bod(:
/ntroduction
0.11 Bac%'round
0.1# 9*Fecti.es I "(pothesis
Methodolo'(
Results
Gimitations
Conclusions and Recommendations.
?.,ppendi4:
Pa'e #$< o- #10
Data collection -orms
Detailed calculations @Computer result
sheetsA
eneral ta*les
Bi*lio'raph(
9ther support materials.
21.+ See Mind8ritter %ritten Re4ort: As an
e0a"4(e: //: 72? , 7+1.
21.. Ora( /resentation: *or"at and
;uide(ines.
9P: , summar( o- the maFor -indin's,
conclusions and recommendations: ,n
opportunit( -or the mana'ers7clients to as%
-or clari-ications and to raise 2uestions.
"int: 9*ser.e the di--erence in se2uencin'.
=hat do (ou !ant the client to %no! at the
end o- presentation

Desired outcomes.
Pa'e #$0 o- #10
9penin', -indin's and conclusions and
recommendations
,daptation to the audience:
E4: Time, lan'ua'e, se2uencin' etc.
,.oid readin' the report and technical
Far'on
Maintain e(e contact
Hse 'raphics and other .isual aids
Pro.ide handouts to the audience.
21.5 Research *o((o8Au4:
Research Reports: Either !ritten and oral -or
decision ma%in'
Research report: &ot the end o- the
research process
Pa'e #$? o- #10
Recontact the decision ma%ers a-ter the(
ha.e had a chance to read the research
report.
Pa'e #$B o- #10
MODULE 1?
ETHICAL ISSUES IN BUSINESS
RESEARCH
22.3 Issues:
=hat are the .arious ethical issues
concernin' -air *usiness dealin's:
/n the perspecti.e o- ethical issues, !hat is
in.ol.ed in the choice o- research techni2ues
as !ell as appropriate use o- research
results:
=hich are the ethical 2uestions relatin' to
parties @sta%e holdersA in.ol.ed in B. R:
=hat are the ri'hts and o*li'ations o- the
researcher, respondent and the compan(
@clientA that sponsors B. R: =hat are the
codes o- *eha.iour -or three parties:
2?.3 Ethics in Research:
Pa'e #$D o- #10
To ensure that no one su--ers ad.erse
conse2uences -rom research acti.ities.
But ho! to ensure:: T!o maFor approaches
Strict adherence to la!
Rel(in' on each indi.idualCs @part(CsA
personal sense o- moralit(
E Some middle 'round:
E Codes and re'ulations 'uide researchers
and sponsors: Research proposal re.ie!
*oards and peer 'roups to help
researchers to e4amine their research
proposals -or ethical dilemmas.
2+.3 ;enera( Ri)hts and O1(i)ations o
/arties in Research:
Pa'e #1$ o- #10
E4hi*it #.1 P.;$
Ethical /ssues and Research Process
enerall( three parties: Researcher, the
sponsorin' client @userA, and the respondent
@su*Fect o- researchA.
2+.1 K Ethica( Treat"ent o Res4ondent
#Su16ect or /artici4ant$: Ri)hts and
O1(i)ations o Res4ondent.
RespondentCs ri'hts E o*li'ation o-
the researcher
E4: RespondentCs ri'ht to pri.ac(


ResearcherCs o*li'ation to protect the
anon(mit( o- the respondent.
: RespondentCs disclosure o-
in-ormation a*out personal matters


Researcher to 'uard that in-ormation
-rom others other than the researcher.
Researcher to -ollo! the -ollo!in'
'uidelines:
o E4plain stud( *ene-its to respondent
Pa'e #11 o- #10
o E4plain respondents ri'hts and
protection @Ri'ht to pri.ac( !hen
a'reed to participate in researchA
o 9*tain in-ormed consent: B( disclosin'
-ull( the purpose o- the stud(, the
respondents o- the stud( etc.
o ,.oid deception: DonCt 'i.e -alse
impression a*out the stud(
9nce in-ormed consent 'i.en, the
o*li'ation to *e truth-ul to 'i.e data.
2+.2 Ri)hts and O1(i)ations o the Researcher:
ResearcherCs o*li'ations to *oth respondents
and clients
Code o- ethics -or researchers:
Ensure o*Fecti.it( so as to pro.ide
accurate results
Pa'e #1# o- #10
E4hi*it #.# P.;+
/n-ormed 5 Consent Procedure -or Sur.e(s
DonCt misrepresent the statistical
accurac( o- their data
Protect the ri'hts to con-identialit( o-
*oth the respondents and clients
E Protection o- ,non(mit(.
DonCt disseminate -ault( conclusions
&ot to use some ideas -rom unsuccess-ul
competin' -irms -or research proFect o- the
sponsor.
E The code o- ethics: De.eloped *(
pro-essional *odies such as Chartered
,ccounts, ,d.ertisin' ,'encies,
Ban%erCs ,ssociation, En'ineerin'
,ssociations etc.
Hse o- computer 5 mediated communication
-or data collection @Data collection in c(*er
spaceA and minin' throu'h tools li%e closed
circuit tele.ision, di'ital camera monitorin':
Man( unsol.ed ethical 2uestions.
Pa'e #1+ o- #10
2+.? Ri)hts and O1(i)ations o S4onsorin)
C(ient:
Ri'ht to recei.e ethicall( conducted research
on !hat e.er the areas such as product,
mar%et, personnel, -inancial etc.
Con-identialit(: Ri'ht to sponsor
nondisclosure, purpose nondisclosure, and
-indin's nondisclosure.
Sponsor &ondisclosure
E4: Testin' a ne! product idea, to enter
into a ne! mar%et !ithout re.ealin' its
plans to competitors.
Purpose &ondisclosure:
E4: Testin' a ne! idea that is (et to *e
patented, in.esti'atin' emplo(ee
complaints, contemplatin' a ne! pu*lic
stoc% o--erin' etc.
>indin' &ondisclosure: Researcher to
maintain con-identialit( o- data and
-indin's.
Pa'e #1; o- #10
Ri'ht to 2ualit( research: Researcher 1not to
lie !ith statistics3.
SponsorCs Ethics:
&ot to as% the researcher to indul'e
in unethical *eha.iour li%e:
o )iolate respondent con-identialit(
o Chan'e data to meet a desired o*Fecti.e
o Chan'e interpretations
o 9mit sections o- data anal(sis and
conclusions
o Recommend *e(ond the scope o- data
collected
Sponsors need education on ethical
.alues o- a researcher
Sponsors to o*ser.e 'eneral *usiness
ethics !hen dealin' !ith researcher
Pa'e #1< o- #10
,.oid misusin' the research -indin's
to support sponsorCs aims
Respect research respondentCs
pri.ac(
Trans-erenc( in intensions to sponsor
research and a*out the *usiness
pro*lem to *e in.esti'ated.
Pa'e #10 o- #10

Você também pode gostar