AttArJtrKl (x'

EAcrm AFmcryc rm AocxlRAtr oil' EAnLy

C{)ST ESIMfrATE IN CI}NSTRUCTK}!{ PROMCTS

IxlnrA$ilArfl I)lvoYolvARsrlo

UNMnSilt

TEKII$OI{)GI ilfiAI.AYSIA

PSZ 19:16(Pind l/07)

UNTVER,SIII IEKNOM}GI MALAYSIA

DECI-A,RATION OF TT{ESIS ANID COPYRIGIIT

Author"s full name

Date ofbirttr

DODI ARDIANTO

:
:

o6ftDEcEwER

1978

Title

ANALYSIS OF FACTORS AFFECTING THE ACCI]RACY OF EARLY COST ESTIMATE IN CONSTRUCTION

PROJECTS

Academic Session: 2010

l20Ll
as:

I declare that this thesis is classified

COI\TFIDENTIAL (Contains confidential information under the Official Secret Act 1972)*

RESTRICTEI)
OPEN

(Contains restricted information as specified by the

organization where research was done)*

ACCESS

I agree that my thesis to be published (tull text)

as online open access

I acknowledged that Universiti Teknologi lvlalaysia reserves the rights

as

follows:

1.

The thesis is the property of Universiti Teknologi Malaysia. The Library of Unive,niti Teknologi Malaysia has the rightto make copies for the purpose of research only. tresis for academic exchange.

2.

3. Ttr Library has tlrc right to make copia

No KP/IC: 201071VI1fi t61

D#19lry2011
Notes :

* if the thesis is COIIFIDENTIAL

restriction

and RESTRICTED, please attach with the letter from the organization with period and reasons for confidentiality or

"I

he,reby

declue thatI have readthisthesis and inmy

opinion this thesis is sufficient in terms of scope and quality for the award
the degre oftdaster ofConsfnrction Contract I\danage, ent

Sigrrdtr Nmrc of Supervisor Date

Signdrc
Name of Second Rader

: PROF.

ffi

Date

'

A}il,IAD ROSDAN RAZAK

I ql'.tt bo/f

JI.JLY 2011

AI{ALY$S OF FACTOR,S AFFECfiNG TTIE ACCI.'RACY OF EARLY COST

ESTIT{ATE IN CONSTRUCTION PROIECTS

DODIARDIANTO

A thesis submitted in ftlfihent of the requimens for thE award offte &gree of I\daster of Soienoe (Cmshrctim Cmnact AdamgmnQ

Facnilty of Built Envirmeffi Univereiti Teknolqgi Ldslaysia

JTJLY 2011

DECLARATION

I declarc that this thesis entitled "Atulysis

in the re&nences. The thesis
has not

Of Faetors Affecting ITu Accuracy

A Early

Cost Ectlmate In Constructl.sn Projects" is the

reult of my onm research except as cited

bm

accepted

fr

any degres and is

nd concumetrly

submitted in candidffirc of any

otrfr

Sigrtfirc

Ihte

Name

: 19 JL]LY2011

tu

dep.

DEDICATION

o'To

my beloved wife, brother, sister, mother, and late father

lv

ACKNOWLEDGMENT

Many thanks to my supervisor, Assoc. Prof. DR. Razali bin Abdul Hami{ for his kind and helpful comments in the preparation of this thesis. Without that, this thesis would not have
been the same as presented here.

I am also thanks to Prof. Ahmad Rosdan Razak for

correction this thesis in order to become more with quality.

I am also indebted to Ministry of Public Works Indonesia for funding my master study. Credit should be given

to all

respondent

for tlreir contributing in

admission filling

questionnaire. My fellow postgraduate students should also be recognized for their support.

My sincere appreciation also

extends to

all my colleagues and others who have provided

assistance at various occasions. Their views and tips are useful indeed.

Finalty, tlnnk you to my beloved parents, brother, sister, my beloved one and whole family
who always support me.

ABSTRACT

Quality early cost estimate can be measured from its level of accuracy. Early cost estimates prepared at the early stage of project inceptions, therefore early cost estimates have an important function as a starting point and a benchmark for project planning and controling in the future project stages. Early cost estimates are prepared based on limited information and supporting data, it couses in early cost estimates has the lowest level of accuracy than any other types of cost estimates.

From the perspective of the important function of an early cost estimate, in contrary the low level of accuracy that is produced is a condition of stark contrast. This condition is often encountered during the preparation of early cost estimates of construction projects that why this topic became the basis of this research. This research aims to make an analysis of the relationship factors that affect the accuracy of the early cost estimate. Research data obtained from surveys of construction projects that have been completed in the period 2006-2011. Survey data were analyzed quantitatively to obtain significant variables and formulate a regression model of factors affecting the accuracy of early cost estimates.

Quality of cost estimates associated with the accuracy and completeness of the supporting elements depends on: nature of the project, the level of data and information available, techniques and methods used, ability of estimator, and other factor considered while preparing the estimate. The the factors were developed into 54 variables as the basic elements of the preparation of early cost estimates. From the results of multiple linear regression analysis obtained significant factors (<0.05) which affects the level of accuracy of the early cost estimates, as follows.: Geographical location / Site location, Quality level of the estimator / Estimator Expertise; Availability project information / Documents used in Preparing the estimate; the availability of scope of the project; the availability of project documents (preliminary and Size of Project / Scope of the project.

ABSTRAK

Kualiti anggaran kos pendahuluan boleh diukur dari tingkat ketepatan yang dihasilkan. Anggaran kos pendahuluan disusun pada peringkat awal projek bermula, sehingga anggaran kos pendahuluan mempunyai fungsi penting sebagai titik tolak dan tolak ukur perancangan dan pengurusan projek pada tahap-tahap berikutnya. Anggaran kos pendahuluan umumnya disusun mengikut maklumat dan data pendukung yang masih terbatas, hal tersebut menyebabkan anggaran kos pendahuluan mempunyai tahap ketepatan yang terendah berbanding jenis anggaran kos lain. Dilihat dari fungsi penting suatu anggaran kos pendahuluan di satu sisi, dan di sisi lain rendahnya tahap ketepatan yang dihasilkan merupakan suatu keadaan yang sangat kontras bertolakbelakang. Keadaan tersebut sering dijumpai pada saat penyusunan anggaran kos pendahuluan projek pembinaan sehingga menjadi dasar dilakukannya kajian ini. Kajian ini bertujuan untuk membuat analisis mengenai hubungan faktor-faktor yang mempengaruhi ketepatan anggaran kos pendahuluan terhadap ketepatan yang dihasilkan. Data penelitian didapati daripada kajian terhadap projek-projek pembinaan bangunan yang telah selesai pada kurun waktu tahun 20062011. Data hasil kajian dianalisis secara kuantitatif untuk mendapatkan variabel signifikan dan merumuskan model regresi terhadap faktor-faktor yang mempengaruhi ketepatan anggaran kos pendahuluan. Kualiti anggaran kos yang berkaitan dengan ketepatan dan kelengkapan unsur-unsur pendukungnya bergantung pada: sifat projek, ketersediaan data dan maklumat, teknik dan kaedah yang digunakan, kualiti dan pengalaman estimator, serta faktor lain yang dipertimbangkan saat menyiapkan anggaran kos. Dari faktorfaktor tersebut dikembangkan menjadi 54 variabel sebagai unsur-unsur dasar penyusunan anggaran kos pendahuluan. Dari hasil analisis regresi linier berganda didapati 4 faktor bermakna (<0.05) yang berpengaruh terhadap ketepatan anggaran kos pendahuluan, iaitu: lokasi projek, kualiti / pengalaman penganggar, ketersediaan data dan maklumat projek, dan lingkup projek.

vi

TABLE OF CONTENTS

CHAPTER

TITLE

PAGE

DECLARATION DEDICATION ACKNOWLEDGMENT ABSTRACT TABLE OF CONTENTS LIST OF FIGURE LIST OF TABLE LIST OF APPENDIX

ii iii iv v vi ix x xi

INTRODUCTION 1.1 1.2 1.3 1.4 1.5 1.6 Background Of The Study Problem Statement Objectives Of The Research Scope Of The Research Important Of The Research Organization Of Thesis

1 1 3 4 4 4 5

LITERATURE REVIEW

vii

2.1 2.2 2.3 2.4 2.5 2.6

Construction Project Project Cost Estimate

7 8

Cost Estimate Classification System and Accuracy Ranges 9 Accuracy Of Early Cost Estimate Early Cost Estimate Factor Affecting Quality of Cost Estimate 15 16 19

RESEARCH METHODOLOGY 3.1 3.3 3.4 Method of Research Research Variables Collection Of Data 3.4.1 3.5 Questionnaire Research

24 24 27 28 29 30 31 31 32 35 35 36 36

Data Analysis 3.5.1 3.5.2 3.5.3 3.5.4 3.5.5 3.5.6 3.5.7 Type Of Data Analysis Correlation Regression Analysis F Test t Test Auto Test Correlation (Durbin Watson Test) Multicolinearity test

ANALYSIS AND DISCUSSION 4.1 4.2 Preparation Research Data Analysis Questionnaires 4.2.1 4.2.2 4.2.3 Analysis Correlation Regression Analysis Model Test

37 37 40 41 47 54

viii

4.2.3.1 Linearity Test ( F Test ) 4.2.3.2 t Test

54 55

4.2.3.3 Autocorrelation Test (Durbin Watson Test) 55 4.2.3.4 Multicollinearity Test 4.3 Result of Research 56 56

CONCLUSION AND RECOMMENDATIONS 5.1 5.2 5.3 5.4 Introduction Conclusion Limitation of the research Recommendation for future research

57 57 57 58 59

REFERENCES

60

ix

LIST OF FIGURE

FIGURE NO

TITLE

PAGE

1.1 2.1 2.2 2.3 2.4 2.5

Typical life cycle profile (Risk Versus Amount At Stake) Project Cost Estimate Process (Oberlander, 2001) The Stages of Project Estimated Cost (Oberlander, 2000) AACE Classification System Design Process Effect of skill and experience of the estimator of the The accuracy of the estimated project costs (Oberlander, 2001)

3 9 9 15 17

20 24 32 34 50 53

3.1 3.2 3.3 4.1 4.2

Research Methodology Correlation Analysis Diagram Regression Analysis Diagram Scatter plot - Regression Standardized Predicted Value Scatter plot - Regression Standardized Predicted Value (end)

LIST OF TABLE

TABLE NO. 2.1 2.2 2.3 3.1 4.1

TITLE Cost Estimate Overview Chart General concept stages of cost estimate process Factor Affecting Accuracy of Early Cost Estimate Likert scale level of accuracy early cost estimate Factor affecting the accuracy of early cost estimate after interview with expert

PAGE 14 16 23 30

40 41 45

4.2 4.3 4.4

Profile of respondents Coeficient Correlation and significant levels of variables Factors that have a significant correlation of early cost estimate accuracy

46 47 47 48 48 49 50 51 51 51 52 52 54

4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16

Model Summary Diagnostic collinearity Coefficient Model Summary 1 Diagnostic collinearity 1 Coefficients 1 Model Summary 2 Diagnostic collinearity 2 Coefficients 2 Model Summary (End) Coefficients (End) ANOVA

xi

LIST OF APPENDIX

FIGURE NO

TITLE

PAGE

A B C D E

Expert Interview Questionnaire Table of F Statistical and t Table Data Questionnaire Summary Respondent

63 70 76 78 82

CHAPTER 1

INTRODUCTION

1.1

Background of the studies Nowadays the construction sector in Indonesia is experiencing rapid growth.

The development of these construction projects marked by numerous large-scale construction projects built by government or private parties. A project has essentially been born from a decision to invest. The decision to make an investment involves a number of funds in hopes to get the benefits that often will have a major impact for the survival of a company. One of the decisions that must be taken to initiate an investment project is to determine what will be built and how much does it cost to build the project. Physical development costs and other expenses are fixed capital for building construction projects. Cost estimate according to the National Estimating Society-USA is the art of estimating (the art of approximating) the possibility of total costs required for an activity of which is based on information available at that time. Based on the function and its relationship to the level of accuracy, there are several types of cost estimates in each stage of the project. For example in the early stages of project concept development, note the estimated how much it cost to build a project, therefore, developed a preliminary cost estimate. Preliminary cost estimate is one important factor that should be considered in decisions to invest in development projects because of the value that result from a process of preparing the preliminary cost estimate is the approximate amount of initial capital needed for the project owner.

2 Evaluation of the feasibility of building projects, and decisions on design and construction issues are usually based on a series of approximate estimates (pre-bid forecasts) that are considered against initial plans and budgets (Betts and Gunner, 1993). Approximate estimates are also required for funding decisions and cost control. Over-estimation or over-provision of funds for one project means lesser funds are available for other business opportunities. In difficult times, estimations of profit margins are likely to be conservative and construction costs are overestimated, otherwise viable projects may be shelved. Estimates form the basis for tender comparison or negotiation, and under-estimation may lead to difficulty in award decisions, or in some cases unrealistic negotiation targets. Projects may be delayed whilst more funding is arranged or even shelved if additional funds are not available. To proceed with insufficient funds will likely lead to payment problems and hefty finance costs for overdrafts or emergency loans during construction. It is therefore necessary for quantity surveyors (QS) to improve the accuracy of their estimates to ensure that clients are satisfied with their services. Accurate, early cost estimates for engineering and construction projects are extremely important to the sponsoring organization and the project team. For the sponsoring organization, early cost estimates are vital for business unit decisions that include strategies for asset development, potential project screening, and resource commitment for further project development. Inaccurate early estimates can lead to lost opportunities, wasted development effort, and lower than expected returns. For the project team, performance and overall project success are often measured by how well the actual cost compares to the early cost estimate. Initial cost estimates are the basis on which all future estimates are compared. Future estimates are often expected to agree with (i.e., be equal to or less than) the initial estimates. However, final cost often exceeds the initial estimate. Estimated project cost is generally a systematic steps and through the same process between one project and previous projects of its kind. Various literatures provide guidance on the process in calculating the project cost estimates using various methods and approaches used. The availability of data and information, time limit, techniques and methods used, skills and abilities an estimator is a factor that determine the quality of the project cost estimate.

3 Another factor affecting the quality of the project cost estimate is uncertainty which is one of the unique characteristics of a construction project. Limitations of data and information in the development stage preliminary cost estimate is one source of uncertainty containing the risk that causes low levels of accuracy. Risk is an unavoidable problem of every construction project. Therefore it is necessary to planning and managing the risk integratedly in every stage of the project, including in the development stage preliminary cost estimate project.

1.2

Problem statement Early cost estimate is a type of cost estimates in the earliest stage of

construction projects. With a high degree of uncertainty (see figure 1.1) because of the limited information available to cause the accuracy of cost estimates has the lowest degree of accuracy than any other cost estimates. But on the other hand the value of projects resulting from the calculation of the preliminary cost estimate is critical to the continuation phase of your next project. That amount will be used as a basis for decision making project implementation and will be used as a reference estimate of the cost at a later stage. Because of the importance of preliminary cost estimation, as a starting point for the commencement of the project and as a basic reference for the next stage, then the accuracy of the preliminary cost estimate is an important thing that should be considered in the process of preparing the preliminary cost estimates.

Figure 1.1 Typical life cycle profile (Risk Versus Amount At Stake) (Source : Budi Manan, 2007)

4 1.3 Objective of the research The objectives to be achieved in this study are: a. Identifying factors that affect the accuracy of early cost estimates of construction projects. b. Identifying relationship between the factor and the accuracy of early cost estimates.

1.4

Scope of the research Scope of the research are: a. Early cost estimates that has been mention in this research is initial cost estimates on construction projects in Jakarta, Indonesia. b. Types of projects to be taken as a sample are building construction project c. The process of preparing the preliminary cost estimates made during the feasibility study lasted until just before the detailed design is completed.

1.5

Important of the research In general, the author hopes this research can contribute to the development

of construction industry in Indonesia, especially in the sector of Construction Management. The important of the research are: a. Providing an explanation of how the availability of data and information, methods and systems used and the experience and skills related to the process estimator estimates the cost could affect the accuracy of preliminary cost estimates. b. Provide input in the application process preliminary cost estimates associated with the things that need to be prepared, techniques and

5 methods used, the necessary procedures and other matters involved in the process preliminary cost estimate in order to produce a preliminary cost estimate is accurate.

1.6

Organization of Thesis

This thesis consists of five chapters, among others:

Chapter I Introduction

In this chapter explained the background of this study, formulation of problem statement , objective of the research, scope of the research, importance of the research, research methodology and organization of thesis.

Chapter II Literature Review

In This section described the basic theory associated with the overall project cost estimates and preliminary cost estimates in particular. Sources literature drawn from several references in the form of books or journals that have published research.

Chapter III

Research Methodology

This section describes the methods and procedures used in this research. Systematic research is illustrated by the flow chart that explains each step that passed from this research. The basis of the steps in quantitative research and data collection system described in this section.

6 Chapter IV Analysis and Discussion

Analysis of data obtained and the discussion until drawn a conclusion described in this section. Every step analysis used and the discussion related to the findings generated are described to obtain a clear picture of the results obtained from this research.

Chapter V Conclusion and Suggestion

This section contains the conclusions that can be drawn from the results of research undertaken. The suggestions that if can be used and further recommendations can be used in the development of subsequent research provided in this section.

CHAPTER 2

LITERATURE REVIEW

2.1.

Construction Projects Projects can be defined as a series of activities that must be implemented with

a limited period, with an allocation of specific funds, and are intended to produce a product or deliverable the quality criteria outlined clearly. Project implementation involves many parties including the owner /owner, Consultants (planners, inspectors, construction management, quantity surveyor), Contractors, subcontractors, suppliers, etc.. With attachment and certain forms of cooperation of each party is working for the implementation of the project. Construction project is one type of project that aims to build a physical building of residential buildings, commercial buildings, industrial buildings. The project has unique characteristics, is always different in terms of complexity, size, and resources necessary. The intensity of the dynamic nature of project activities, the dynamics is characterized by ups and downs of resource requirements throughout the project cycle. Because of the complexity and dynamics of the intensity of the project, the project is divided into several stages. There is no standard that is used in the division of stages in a project cycle, this staging is generally based on the type of activities that take place during these stages. PMI (Project Management Institute) Divide the project into phases: Conceptual Stage, Stage Development Planning (PP / Definition), Implementation and Termination. The purpose of the division of stages in one cycle of this project is to facilitate in identifying and tracking changes in the activities so that project management system can be effectively done.

8 The initial phase of the project begins with the conceptual development plan project. This stage is considered a very important stage in the project cycle, because in this stage that the decision to begin or whether a project or investment is made. At this stage determined some important things like the development of initial plans of the basic concepts and outline the project scope, plan product capacity, financing plan, plan schedules, determining the location, etc. Therefore this stage is the stage which has a major influence in determining the success of a project in terms of quality, cost and project schedule. Over the project when implementation of the project has been running, which has defined the project scope, budget and project implementation has been assigned the project completion time has been determined, it will be very difficult and it will be too late to make significant changes to improve quality, optimize cost and schedule changes project.

2.2.

Project Cost Estimate Estimating project cost is an important element in managing the project,

because one of the parameters of project success is considered from the aspect of cost performance. Basically, the project cost estimate done to find out how much it cost to build a project or investment. Furthermore, cost estimates have a function with a very broad spectrum, namely planning and controlling resources. National Estimating Society-USA defined estimated cost cost estimate as the art of approximating possibility of total costs required for an activity based on information available at that time. From the above definition, the process of preparing an estimate of project costs associated with activities to analyze costs based on available information and estimated costs to be and may occur in the future. Cost analysis and discussion focuses on the assessment of costs based on information available to our times and the past. While cost estimates are meant to see the future taking into account and make an estimate of the things that will and might happen in the future.

9 Estimating project cost is generally a step-by-step through the process of systematic and relatively the same between one project and previous projects of its kind. Project cost estimate process depicted in Figure 2.1 below:

Figure 2.1 Project Cost Estimate Process Source : (Oberlander, 2001)

2.3.

Cost Estimate Classification System and Accuracy Ranges The process of cost estimates made during the project cycle, some types of

cost estimates are developed based on the stage in a project cycle (Figure 2.2), beginning with preliminary cost estimates at this early stage conceptual continued to expand in the next stage until all the stages of construction in progress (Oberlander, 2000)

Figure 2.2 The Stages of Project Estimated Cost Source : (Oberlander, 2000)

10 A variety of terms used by some literature to classify and define the type of project cost estimates. Based on its function and the stages through which as well as supporting elements, Barrie (1992) classify the types of cost estimates as follows: a) Preliminary Cost Estimates / Conceptual and Preliminary Estimate. Preliminary cost estimate are cost estimates prepared in the early stages of planning the project before the completion of engineering design is made. b) Estimated Cost Details / Detailed Estimate Detailed cost estimates are prepared cost estimates based on planning documents and technical specifications of engineering that has been made. c) Definitive Cost Estimates / Definitive Estimate The estimated cost is definitive cost estimates used to predict the cost of the project (The project cost forecasting) Based on constraints derived from the estimated cost of which was made earlier. The American National Standards Institute (ANSI, 1991) defines three types of estimates: order-of-magnitude, budget, and definitive. a. Order-of-magnitude estimates have an expected accuracy between +50% and -30%. They are generally based on cost-capacity curves and costcapacity ratios and do not require any preliminary design work. b. Budget estimates are based on flowsheets, layouts, and preliminary equipment descriptions and specifications and have an accuracy range of +30% to -15%. Design generally must be 5 to 20% complete to permit such an estimate to be performed. c. Definitive estimates require defined engineering data, such as site data, specifications, basic drawings, detailed sketches, and equipment quotations. Design is generally 20 to 100% complete, and estimate accuracy should be within +15% to -5%. Definitions of these estimate types were developed by AACE International. AACE International has proposed an expansion of the ANSI estimate

11 classifications to five types with expected accuracy levels based upon the amount of project definition available when the estimate is prepared (AACE, 1997). The accuracy of each class of estimate depends on the technological complexity of the project, appropriate reference information, and inclusion of an appropriate contingency determination. In all cases, accuracy ranges could exceed the ranges indicated below in unusual circumstances. The revised classifications are:
a.

Class 5 estimates: These estimates are generally based on very limited information. They may be prepared within a very limited amount of time and with little effort expendedsometimes less than one hour. Often little more is known than the proposed plant type, location, and capacity. This class of estimate falls into the ANSI order-of-magnitude classification. The required level of project definition is 2% or less and the expected accuracy is -20% to -50% on the low side and +30% to +100% on the high side.

b.

Class 4 estimates: Class 4 estimates also are generally prepared based upon limited information and also have fairly wide accuracy ranges. They are typically used for project screening, feasibility determinations, concept evaluation, and preliminary budget approval. Engineering is only 1% to 5% complete and comprises, at a minimum, plant capacity, block schematics, indicated plant layout, process flow diagrams (PFDs) for the main process systems, and preliminary lists of engineered process and utility equipment. Typical accuracy ranges for this class of estimate are -15% to -30% on the low side, and +20% to +50% on the high side. This class of estimate falls into the ANSI budget estimate classification.

c.

Class 3 estimates: These are estimates which form the basis for budget authorization, appropriation, and/or funding. These estimates

typically form the initial control estimate against which all actual costs and resources will be monitored. The required level of project definition (i.e., completed engineering) is 10% to 40% and

12 includes at a minimum: process flow diagrams, utility flow diagrams, preliminary piping and instrument diagrams, plot plans, developed layout drawings, and essentially complete engineering process and utility equipment lists. Accuracy ranges for this class of estimate are -10% to -20% on the low side, and +10% to +30% on the high side. This class of estimate also falls into the ANSI budget estimate classification.
d.

Class 2 estimates: This class of estimate falls into the ANSI definitive estimate category. Class 2 estimates are generally prepared to form detailed control baselines against which all project work is monitored in terms of cost and progress control. For contractors, this class of estimate is often used as the bid estimate. Typically engineering is 30% to 70% complete and comprises, at a minimum: process flow diagrams, utility flow diagrams, piping and instrument diagrams (P&IDs), heat and material balances, final plot plans, final layout drawings, complete lists of engineered process and utility equipment, single line electrical diagrams, electrical equipment and motor schedules, vendor quotations, detailed project execution plans, resourcing and work force plans, etc. Accuracy ranges are much improved over the prior classes of estimates. On the low side they are 5% to -15%. On the high side, the ranges are +5% to +20%.

e.

Class 1 estimates: Also included in the ANSI definitive estimate category, this is the most accurate classification of estimates. Class 1 estimates are generally prepared for discrete parts or sections of the total project rather than for the entire project. The parts of the project estimated at this level of detail are typically used by subcontractors for bids, or by owners for check estimates. The updated estimate is often referred to as the current control estimate and becomes the new baseline for cost/schedule control of the project. This type of estimate is often made to evaluate and/or dispute claims. Typically engineering is 50% to 100% complete and comprises virtually all engineering and design documentation of the projects, and complete project

13 execution and commissioning plans. Typical accuracy ranges are 3% to -10% on the low side and +3% to +15% on the high side. Another associated cost estimating reporting and formatting reference document is the ASTM E2516 Standard Classification for Cost Estimate Classification System for Building Projects, General Construction Projects, and Infrastructure Projects. In general, this system is intended to be used for projects where the main purpose of the project pertains to a building or facility. This shall also include general site construction and infrastructure projects, where a building or infrastructure project is the majority of the overall scope of the project. Table 2-1 below is an overview chart that combines tables from AACE 18R-97 and ASTM E2516 Cost Estimating Classification System (see recommended practices for more information):

14
Table 2-1

Cost Estimate Overview Chart

Primary Characteristic LEVEL OF PROJECT DEFINIITION Secondary Characteristic END USAGE METHODOLOGY PREPARATION EFFORT [b] EXPECTED ACCURACY RANGE [a]

Estimate Class

Expressed as % of Complete Definition

Typical Purpose of Estimate

Typical Estimating Method

Typical degree of effort relative to least cost index of 1 [b]

Process Industry (AACE 18R-97)

Building and General Construction Industry (ASTM E2516)

Class 5

0% to 2%

Class 4

1% to 15%

Capacity Factoring, Concept Screening Parametric Models, or Feasibility Judgment, or Analogy Equipment Factored, Parametric Models, , Assembly Drive Concept Study or Models Feasibility Semi-Detailed Unit Costs with Assembly Level Line Items Detailed Unit Cost with Forced Detailed TakeControl or off Bid/Tender Check Estimate or Detailed Unit Cost with Bid/Tender Detailed Take-off Budget, Authorization or Control

Low: -20% to -50% High: +30% to 100%

Low: -20% to - 30% High: +30% to +50%

2 to 4

Low: -15% to -30% High: +20% to 50% Low: -10% to 20% High: +10% to 30% Low: -5% to -15% High: +5% to +20% Low: -3% to -10% High: +3% to +15%

Low -10% to -20% High: +20% to +30% Low: -5% to -15% High: +10% to +20% Low: -5% to -10% High: +5% to +15% Low: -3% to -5% High: +3% to +10%

Class 3

10% to 40%

3 to 10

Class 2 Class 1
Notes:

30% to 70% Up to 100%

4 to 20 5 to 100

[a]

The state of proecess technogy and availibility of applicable reference cost data affect the range markedly. The +/- value reprsents typical percentage variation of actual cost from the cost estimate after application of the contingency (typically at a 50% level of confidence) for given scope. If the range index value of '1' repesents 0.005% of project costs, then an index value of 100 represents .05%. Estimate preparation effort is highly dependent upon the size of the project and the quality of estimating data and tools.

[b]

Sources:

1. 2.

AACE International Recommended Practice 18R-97 Cost Estimate Classification System - As Applied in Engineering, Procurement, and Construction For The Process Industry, Morganton, VA, 1997 ASTM, Standard Classification For Cost Estimate Classification System, E2516-06, West Conshohocken, PA, 2006

15 Figure 2.3., shows the various ranges of the AACE 18R-97 Recommended Practice Cost Estimate Classification System As Applied In Engineering, Procurement, and Construction For The Process Industries:

Source : AACE International Recommended Practice

2.4.

Accuracy Of Early Cost Estimate Early stage cost estimation is the forecasting of the cost of a project during

the planning and design stage (Serpell, 2005). Skitmore (1991) describes the accuracy of early stage estimation as comprising two aspects, namely, bias and consistency of the estimate when compared with the contract or accepted tender price. Bias is concerned with the average of differences between actual tender price and estimate while consistency of estimates is concerned with the degree of variation around the average. Aibunu and Pasco (2008) describes bias (inaccuracies) in the estimate of a project may arise from two sources, namely, bias associated with the project itself (will be the same regardless of the estimator) and bias associated with the estimating technique used and the environment (which would change depending on the estimator).

16 2.5. Early Cost Estimate Estimated project cost is a dynamic activity conducted throughout the project cycle stages, starting from the beginning of the project and ends when the projects are handed over to the project owner (Gould, 2009). Viewed from the process in the project cycle, early cost estimates are cost estimates prepared during the feasibility study phase until just before detail design has been completed (Oberlander, 2000). So basically the cost estimates are preliminary cost estimates are compiled based on data and information from an initial concept in the form of an outline of the scope of a project. Various terms of some of the literature used to explain the preliminary cost estimates, such as: early cost estimate (Oberlander, 2000), Conceptual estimate (CMAA, 2001), Preliminary estimate, parametric estimate (Hegazy 2002), Feasibility estimate, an order of magnitude estimate (Kerzner, 2005), screening estimate (Ritz, 1994), pre-tender estimate (Ashworth) etc.. Preliminary Estimates are employed in the early planning phases of a proposed project to match an owner's needs (see table 2.2 ), expressed as written requirements, with budget constraints in order to establish its overall scope (size) and quality expectations. (Rosdan, 2010)

Table 2.2 General concept stages of cost estimate process Source : (Rosdan, 2010)

17 Prelininary Estimates based only on guide prices : (Unit, Superficial, Cube, Storey enclosure, Approximate quantities, Resource analysis, Cost models. (Rosdan, 2010) Early stages at project development clearly define in figure 2.4 below :

Figure 2.4 Design process Source : (Rosdan, 2010) In the early stages of the project, a concept will be formulated in the form of a project outline project scope and conceptual design and preliminary project development programs. Based on the data and information that form the initial concept of a project, preliminary cost estimates have been prepared to give an idea of how big the budget should be spent to make it happen and provide guidance in project development at a later stage. Under conditions limited data and information, the biggest challenge in preparing the preliminary cost estimate is to interpret the data and information into a value of the cost of a project. Furthermore, based on the results obtained from the preliminary cost estimates and with a certain accuracy to decided level, conceptual budget (conceptual budget) Project assigned. This conceptual budget will then be used as a basis for consideration in determining the strategy to be applied to subsequent stages.

18 Preliminary cost estimates is part of a project feasibility study whose results will be used as one basis in decisions about the steps that will be done next, for example in terms of planning and development of corporate strategy, screening of projects wherever potential harm to the company's business continuity and determination of policy in terms of the use of company resources. In terms of availability of budget funds that are owned by the owner of the project and how much budget is needed to build the project, the results obtained from the preliminary cost estimates will be used as a reference for comparison against the budget funds that are owned by the project owner. So the preliminary cost estimate is kind of the first cost estimates produced in a project cycle which made reference to check if the budget funds /budget in accordance with the cost needed to realize the project (Barrie, 1992). When assessed the cost is too great to budget funds /budget available (overbudget), Then the alternative concept will be developed to suit the available resources. At this stage of the Planning and Development (PP / Definition), preliminary cost estimates used by the project team as one of the considerations in formulating strategy and management of projects as the basis of stats and technical design development projects. Preliminary cost estimates are also used as a basis for preparing the outline of the project schedule (Barrie, 1992). In the outline of the schedule is set to determine when the planning phase of design and implementation procurement can be implemented. Usefulness of the preliminary cost estimates by Iman Soeharto (2001) is as follows: a. b. c. Assessing the economic and financial feasibility of the project. Determining the order of priority (ranking) Of several projects. Determining whether continued or not efforts further assess the feasibility of the project. From the above explanation can be concluded on the importance of a preliminary cost estimate, but the limited data and information available is one of the

19 causes of type preliminary cost estimates have the lowest degree of accuracy than any other cost estimates. The more complete information available, the better the resulting accuracy. Therefore it needs to do the right approach in preparing the preliminary cost estimates to generate cost estimates that have a good degree of accuracy.

2.6.

Factor Affecting Quality of Cost Estimate Skitmore and Stradling (1990) describes five primary factors affecting quality

of cost estimating. These concern (1) the nature of the project, (2) the information used, (3) the estimating technique used, (4) the feedback mechanism used and, (5) the person providing the forecasts. This section provides a review of the literature that has been examined to date on the subject. Similar with Skitmore, Iman Soeharto (2001) describes the quality of cost estimates depend on the following matters: a. b. c. d. Availability of data and information Techniques and methods used Proficiency and experience of the estimator. The purpose of the use of cost estimates.

The availability of information about the scope of the project is an important factor in the quality of the resulting cost estimates. According to Oberlander (2000), the accuracy of the estimated project cost is a function of the availability of information (definition of project scope). The further progress of the project, estimated cost or budget, the better prepared and sharp accuracy, because the availability of input data and information needed (Iman Soeharto, 2001). For example, in the early formulation of the scope of the project, because most of the data and information is not available or can not be determined, prepared a preliminary cost estimate with accuracy that is still rough (order of magnitude.) Therefore, the increasing availability of data and information along with the progress of the project are known several types of costs during the project cycle.

20 Techniques and methods used affect the quality of the resulting cost estimates. However, the selection of techniques and methods do not stand alone, but closely related to the purpose of the use of cost estimates and information available. Due to the nature of work within the estimated cost of requiring various assessment and judgment especially in the preparation of preliminary cost estimates where data and information available is still limited, the skill and experience compiler cost estimate (estimator) is very influential to produce quality cost estimate. Figure 2.5 illustrates the influence of skills and experience of the estimator, procedures and methods used to estimate the accuracy of a cost.

Figure 2.5 Effect of skill and experience of the estimator of the accuracy of the estimated project costs Source : (Oberlander, 2001)

In one known term project cycle cost of the project cycle (cost engineering) which includes the cost estimates, cost budgeting, and cost control. So a person who served in one cycle cost engineering must have experienced, because the cost aspect is very important benchmark in the success of the project. Besides the above factors, the quality of cost estimates is also determined by the purpose of use. For example, preliminary cost estimates used in assessing the feasibility of the project does not need to be prepared outlining the scope of work

21 (work breakdown structure) To detail as in the process of developing a definitive cost estimates used in preparing the project budget control. Based on the explanation and reviews from several literature above, the authors make a grouping of factors that affecting the accuracy of early cost estimate into 5 groups:

No. I. 1. 2. 3. 4. 5. 6. 7. II. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

VARIABLES NATURE OF THE PROJECT Type of the project Size of project / scope of project Complexity of project Other project characteristics Geographical location / Site location The contract procurement system The prevailing economic climate

LITERATURE

Skitmore (1999); Ogunlana and Thorpe (1991) Skitmore (1999; Ogunlana and Thorpe (1991) Ogunlana and Thorpe (1991) Skitmore (1990) Skitmore (1990); Ogunlana and Thorpe (1991); Oberlander (2003) Skitmore (1990) Skitmore (1990)

THE LEVEL OF DATA AND INFORMATION AVAILABLE Completeness of cost information Quality of information Accuracy and reliability of cost information Aplicability of cost information Project information/Documents used in preparing the estimate Capacities of project Time allowed for preparing the estimate Technology; Method of construction Processes project Piping & Instrumentation Diagrams Amount of special work likely production time Project design criteria Complexity of design and construction Utility resources and supply conditions Heat and material balances Oberlander (2003) Babalola (2006) Oberlander (2003) Oberlander (2003) Babalola (2006) Oberlander (2003); Babalola (2006) Oberlander (2003) Oberlander (2003); Babalola (2006) Oberlander (2003) Oberlander (2003) Babalola (2006) Oberlander (2003) Babalola (2006); Ogunlana and Thorpe (1991) Oberlander (2003) Oberlander (2003)

22
23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. III. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. Mechanical equipment list Lead time for delivery of materials Project Schedule The form of procurement and contractual aggrement Project strategy The extent of completion of pre-contract decision Frequency of construction variations Clients financial situation Owners cost Purpose and intended use of estimate Owners experience level Engineer/Designers experience level Enviromental assesment and workforce The type of structure Oberlander (2003) Babalola (2006) Oberlander (2003); Babalola (2006) Babalola (2006) Oberlander (2003) Babalola (2006) Babalola (2006) Babalola (2006) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003); Babalola (2006) Babalola (2006)

THE ESTIMATING TECHNIQUE AND METHOD USED Alligment of estimate methodology with available project information Is the estimate work process formally define and followed? Formal structure to categorized and prepare the estimate Utilization of checklist to ensure completeness and technical basis Standard procedure to update cost data of project Documentation of information used in preparing the estimate Involvement of other resources in preparing the estimates Level of involvement of the project manager Level of team integration and aligment Review and acceptance of estimate by appropriate parties Attitude/culture forward changes Method used to determine contingency the conference estimate, comparison estimate, graphical relationships, unit techniques, functional approach, resource estimate Oberlander (2003) Oberlander (2003) Oberlander (2003) Babalola (2006) Oberlander (2003); Babalola (2006) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003); Babalola (2006) Ashworth and Skitmore (1986) Ashworth and Skitmore (1986) Ashworth and Skitmore (1986) Ashworth and Skitmore (1986) Ashworth and Skitmore (1986) Ashworth and Skitmore (1986)

23
55. 56 IV. 57. 58. 59. 60. 61. V. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. factor estimate exponent estimate ABILITY OF ESTIMATOR Estimator expertise Relevant experience of the estimating team Quality level of Estimator Attributes of Estimator Acquisition and application of expertise Iman Soeharto (2001) ; Babalola (2006) Oberlander (2003); Ogunlana and Thorpe (1991) Skitmore (1990) Skitmore (1990) Skitmore (1990) Ashworth and Skitmore (1986) Ashworth and Skitmore (1986)

OTHER FACTOR CONSIDERED WHILE PREPARING THE ESTIMATE Impact of project type Impact of contract type Impact of project schedule Impact of govermental requirements Work force Labour productivity Tax and insurence Money factors (currency, inflation, Bidding climate The nature of the competition / Bidding Competitiveness Logistic for engineering and constructions Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003) Oberlander (2003); Babalola (2006) Oberlander (2003) Oberlander (2003) Skitmore (1990); Babalola (2006) Oberlander (2003)

Table 2.3. Factor Affecting Accuracy of Early Cost Estimate Source : (Skitmore, Ashworth, Ogunlana, Thorpe, Oberlander, Babalola, Iman Suharto)

24

CHAPTER 3

RESEARCH METHODOLOGY

3.1.

Method Of Research

In general, the research methodology is a step-by-step or series of activities starting from the beginning of the study until the final report of a study. The research method is indispensable in providing guidelines for researchers to achieve the goals and objectives of the study. The research method is closely related to procedures, tools and research designs used. Research design, procedures and equipment used in research must comply with the research method used. Based on the rules in scientific research, following the steps used in this study (Figure 1.1).

Figure 3.1: Research Methodology Source : (Nazir, 2005)

25 In this research method used is survey method, the research conducted to obtain the facts from the existing symptoms and seek factual descriptions, whether of social institutions, economics, or politics of a group or region. Survey methods to dissect and skinning as well as recognize the problems and get the justification of the status and practices that are in progress. (Nazir, 2005). Research conducted at the same time against a number of individuals or units using sample.

The first step in doing research is to formulate the problem to be solved. To remove any doubt the matter should be clearly identified, namely by defining the extent the scope and limitations of the problem and research objectives to be achieved. Scope / limitations of the problem and research objectives should be consistent with the formulation and definition of the problem.

After the formulation of the problem, then proceed by searching for data related to the topic of research. Study literature conducted to gain a deeper understanding on the aspects studied. The data obtained from books and research journals and the search for information via the Internet. In the field of science that already have strong theories, it is necessary to formulate a theoretical framework or conceptual framework which can then be derived in the form of hypotheses to be tested.

After the data obtained in accordance with the problems to be solved, the next step is to formulate a research hypothesis. The hypothesis is tentative conclusions about the relationship between variables has to do anything to or phenomena in the study. The hypothesis is a tentative conclusion that acceptable while before the test. After the hypothesis set, the next step is to formulate ways to test the hypothesis. Hypothesis testing based on a framework of analysis (analytical framework) which has been set.

The hypothesis test requires data, the data can be primary data and secondary data related to research problems. The data collected is in the form of facts that will be used to test the hypothesis that has been formulated previously. Data collection techniques vary depending on the problem to be solved and the research techniques

26 used. In this research a survey involving interviews of experts who are experts in cost engineering. It is important to validate the data obtained through the study of literature, so in accordance with the conditions of construction projects in Indonesia. Then, based on information obtained from the literature and input from experts, developed a questionnaire as a tool for collecting research data.

After the data collected, the data is arranged in advance to facilitate analysis. Compilation of data can be in the form of tables and make coding which will then be analyzed with the help of computer software. In this study, an analysis aimed at identifying risk factors significant happens during the preparation of cost estimates that affect the accuracy of preliminary cost estimates. A mathematical model of the analytical results generated to reflect the relationship between phenomena that are implicit contained in the hypothesis, which will then be tested by several statistical techniques.

The analysis result obtained then interpreted and made generalizations based on findings that there is, so few conclusions can be drawn. This conclusion should be related to the research hypothesis that has been formulated previously. What is the correct hypothesis to be accepted, or whether the hypothesis was rejected. What is the relationship between the phenomena obtained will be applied generally or only apply to the particular conditions only. What suggestions can be drawn from the research and how their implications for adoption.

After the formulation of the problem, then proceed by searching for data related to the topic of research. Study literature conducted to gain a deeper understanding on the aspects studied. The data obtained from books, research journals and other source that related to the topic of research.

Study literature conducted to gain a deeper understanding on the aspects studied. In the study of theories of literature related to research topics collected. This is done to obtain a broader orientation on the problems to be solved, and avoiding duplication is undesirable. Aside from being a source of secondary data, literature

27 studies are also needed to know to which science-related research topics and to develop where there are conclusions and degeralisasi ever made before.

Studies on the research literature discusses several topics related to the early cost estimates of construction projects. The estimated cost of the project outline discussed from the start definition, the process is done until the type and function of the estimated project cost. In this section, early cost estimates discussed in more depth, with respect to the functions and processes that must be passed. Explained about some essential functions of a preliminary project cost estimates as cost estimates were first made at an early stage where a new project starts.

3.2.

Research Variables

After the research problem and research hypotheses are formulated and carried out literature studies, further determined what variables to use in these research. The variables that you want to use should be established, identified and classified in accordance with the desired goals and objectives achieved (Nazir, 2005).

In this research, there are two research variable, is described as follows: a. The independent variable or independent variable (Xi), ie the risk factors that affect the accuracy of preliminary cost estimates, with i = 1,2,3, ... b. Dependent variable or dependent variable (Y), namely the variable measuring the level of accuracy of the preliminary cost estimates on construction projects that are influenced by the variable X.

The variables of this research are the factors that affecting the accuracy of early cost estimates. These factors were collected from various sources of literature are used as secondary data used in evaluating the quality of early cost estimates associated with the resulting level of accuracy. Research variables can see in chapter 2 table 2.2 Factor Affecting Accuracy of Early Cost Estimate

28 Research variables derived from a concept that needs to be clarified and changed its shape so it can be measured and used operationally. Proper gauge for measuring the variable or concept is very important. With the appropriate measuring instruments, researchers can link an abstract concept to reality and to formulate and test hypotheses without obtaining difficulty (Nazir, 2005). By using the size / scale of the right of a concept or variable is qualitative, the concept or variable will have a quantitative trait. In other words, the scale needed to change the attribute with the qualitative characteristics into a quantitative variable.

3.4.

Collection of Data

Data collections here mean a process of procurement of primary data and secondary data for research purposes. Data collection is a very important step in the scientific method, because the data collected is used to test the hypothesis that has been formulated previously.

Variables obtained from the study of research literature to be seen again one by one its relevance to the limits and scope of the research. Often these variables can not be used because it does not comply with the limits that have been made previously. Duplication of variables that mean the same should be avoided to get good research data.

Interviews are one of the data collection techniques in the research process. Data collected must be sufficiently valid to be used. In this study, variables derived from prior literature study verified by experts through interviews with experts who are competent, both from the practitioners and academicians. Interviews with experts committed with intent to obtain or confirm a fact as well as to increase confidence about the state of facts. Through interviews with experts, secondary data in the form of risk variables that affect the accuracy of preliminary cost estimates obtained from literature will be reduced or even be increasing in number. This is done to identify the variables may not be found in the literature but significant happens in practice.

29 3.4.1 Questionnaire Research

Technique of primary data collection in this research is by distributing a list of frequently called a questionnaire. The questionnaire contains questions based on variables obtained through literature studies and inputs from experts who have done previously. The questionnaire was divided into two parts, the first part contains data on the background of the respondents and the second part is the core of the questionnaire which contains the factors that affect the accuracy of preliminary cost estimates.

In the survey questionnaire design, independent variables are arranged in a list of questions about how the factors that affects the accuracy of preliminary cost estimates that occurred in the project of the respondents. Answer the question made in the form of options that have been prepared on a Likert scale with 5 (five) options.

The dependent variable in this study is the accuracy of preliminary cost estimates as measured by comparing the cost estimates generated in the process of project cost estimates with actual costs of the project, namely the costs incurred to build the project / cost realization. The accuracy of preliminary cost estimates can be measured in the formula:

Br - Be Y = ----------------- x 100% Be

Where:

Y = The accuracy of the preliminary project cost estimate Br = Actual Cost Be = Cost Estimates (Source : Ismed, 2008)

30 Choice accuracy was measured in the early cost estimate following scale: Scale 1 2 3 4 5 Quantitative Assessment - 50% Y - 30% or + 50% Y + 100% - 20% Y - 10 % or + 20% Y + 50% -10 % Y - 5 % or +10% Y + 20% -5 % Y - 3 % or +5 % Y + 10% -3 % Y + 5% Level Accuracy Inaccurate Less accurate Fairly accurate Accurate Very Accurate

Table 3.1

Likert scale level of accuracy early cost estimate

(Source : Analysis Result from AACE 18R-97 and ASTM E2516)

3.5.

Data Analysis

From the data obtained and collated, it is necessary to look for patterns of data analysis these factors. The analysis used should be a proper analysis of existing data to manage, so the results are consistent with the objectives to be achieved.

The process of further analysis is to use the software help Microsoft Exel and Statistical Product for Service Solution (SPSS). This analysis was conducted to process data that has been distributed questionnaires to the respondents. SPSS software is a computer application programs which are specifically used to analyze statistical data. This analysis is done to see how much influence the risk variables on the accuracy of preliminary cost estimates, which in turn acquired modeled relationship model. Then do the test model in the form of quantitative analysis: statistical analysis, correlation analysis, regression analysis, testing

heteroscedasticity, multicollinearity test, t test and F test.

31 3.5.1 Type of Data The data to be used in the analysis of the questionnaire is interval data. Interval data (also sometimes called integer) is measured along a scale in which each position is equidistant from one another. This allows for the distance between two pairs to be equivalent in some way. This is often used in psychological experiments that measure attributes along an arbitrary scale between two extremes. In this case, researcher determine level of accuracy, rate from 1 to 5 as a data. (ref : changingminds.org)

3.5.2

Analysis Correlation Correlation analysis is a numerical relationship between two random

variables. Correlation analysis aims to identify and find the relationship between several variables that have been established for research. In the correlation test will test how close the relationship between free variables and the dependent variable. Correlation does not show any relationship, but merely gives an indication of strength of the relationship between the variables.

The relationship of two variables declared with correlation coefficient (r). The direction of the relationship between two variables can be divided into: a. Direct correlation (positive correlation), namely a change in one variable followed other changes regularly with the same direction of motion. b. Inverse correlation (negative correlation), namely a change in one variable followed by changes in other variables on a regular basis with movements in the opposite direction. c. Uncorrelated (No Correlation)

(1) Positive Correlation

(2) Negative Correlation

(3)Uncorrelated (No Correlation)

32 The relationship between variables produce positive or negative value with the limit value of the coefficient correlation r (Pearson Correlation Coefficient) is 1 for positive correlation and -1 for negative correlation. If the value of correlation coefficient close to zero, meaning there is no linear relationship between variables.

Diagram step of correlation analysis is as follows: Correlate

Bivariate

OK

Correlation Coefficients : Pearson Test of significant : Two-tailed Flag significant correlation Options: Statistics : Means and standard deviations Cross-product deviation and covariences Missing value : Exclude cases pairwise

Input all variabel include accuracy into thevariables box .

Figure 3.2 Correlation Analysis Diagram

3.5.3

Regression Analysis

Correlation analysis is used to determine if there is a relationship between two or more variables, while regression analysis is useful to predict how far the influence of one or several independent variables on the dependent variable. Regression analysis is one very important statistical analysis and mathematical modeling related to the problem of a pair of observation data. Regression is a statistical technique to determine the equation of a line or curve by minimizing the deviation between observation data and the value of the regression equation. The purpose of regression analysis in general is: a. Determining the equation of the regression line based on the rate constants and the resulting regression coefficients. b. Finding correlations jointly between free and bound variables. c. Testing the significance of independent variables affect the dependent variable.

33 Regression model generated from this regression analysis show quantitatively the relationship between free variables X and Y, as follows:

Y = a + bX

a=

( Yi )( X i ) ( X i )( X i Yi )
2

n X i ( X i )
2

b=

n X i Yi - ( X i )( Yi ) n X i ( X i )
2 2

(Sudjana 2005: 315). Where:

a = Constant b = Regression coefficient

X = Independent variable Y = Dependent Variable

This study uses multiple regression analysis that is a linear regression analysis consists of a dependent variable and several independent variables. Multiple linear regression analysis indicated the relationship between the regression equation:

= a + a X + a X + ... + a X Y 0 1 1 2 2 k k
a0 = Y a1 X 1 a2 X 2

a1 =

( X 2i )( X 1iYi ) ( X 1i X 2i )( X 2iYi )
2

( X 1i )( X 2i ) ( X 1i X 2i )
2 2 2

a2 =

( X 1i )( X 2iYi ) ( X 1i X 2i )( X 1iYi ) ( X 1i )( X 2i ) ( X 1i X 2i ) 2
2 2

(Sudjana 2005: 349).

34

Where : X 1 , X 2 , ..., X k = Independent variable Y a0 a1 , a 2 ,..., a k = Dependent variable = Constanta = Regression coefficient

Diagram step of regression analysis is as follows: Analyze d Regression Linear

Linear Regression : Input variabel dependent Input variabel independents Method : Enter Input Statistics Input Plots

OK

Figure 3.3. Regression Analysis Diagram

Regression model that has been obtained from regression analysis of model validation should be qualified as follows: 1. Value Adjusted Square (R2) To measure the relationship between independent variable X on variable Y has a minimum value of 0.5. 2. Value multi-collinearity has a maximum value of 18. 3. Regression coefficients must all be positive

In multiple regressions analysis is expected between the free variable (X) there is no strong correlation because if this happens then the variable should not be put into the equation. The existence of a strong correlation can be seen from the VIF value > 10. If VIF < 10, then there is no strong correlations (multicolinearity).

35 In practice, multiple regressions is more widely used because of the many variables that need to be analyzed together, and also are generally more relevant regression is used.

3.5.4

F test

F test used to test the null hypothesis (H 0 ) that all coefficients of independent variables Xi of the regression model is zero, and the alternative hypothesis is (H a ) is that the entire value of the coefficient of variable X is not equal to zero. In other words the ratio of F used to test the null hypothesis (H 0 ), namely that free variables together is not affected on the dependent variable, and alternative (H a ), namely that the independent variables affect the dependent variable tehadap. Good research data tend to come from the cause of the smaller of the result, so there is clarity of contribution to the impacts generated. Variance is the difference of each data with the average.

3.5.5

t test

t test used to test the null hypothesis (H 0 ) that each coefficient of the regression model is zero and the alternative hypothesis (H a ) is that if each coefficient of the model is not equal to zero. If the null hypothesis is accepted means that the resulting model can not be used to predict Y, otherwise if the hypothesis is rejected, then the value of the resulting model can be used to predict the value of Y. Criteria for testing this hypothesis is as follows: H 0 is rejected if t 0 count > t a (n-k 1 ) table H 0 is accepted if t 0 count < t a (n-k -1 ) table

36 3.5.6 Auto Test Correlation (Durbin Watson Test)

Dubin Watson test was conducted to examine there is or there is no auto correlation between the variables studied. Test the auto correlation with the value restriction Durbin Watson (-2 < X < +2) to determine the presence or absence of residual correlation or auto correlation of the resulting regression model.

3.5.7

Multicolinearity test

Multicollinearity is a condition in which independent variables in regression equations have significant correlation with the other, or in other words there is a correlation among independent variables (variables x). Parameters that can be used as a benchmark whether or not multicollinearity is: (a) Look at the correlation between pairs of variables: if the correlation coefficients are greater than 0.80 (sometimes 0.90), the variables are strongly correlated and should not be used. (b) Use indicators that are calculated by SPSS Tolerance The percentage of variance in a variable not associated with other variables. Tolerance has a range from zero to one. A value of near 1 indicates independence. If the tolerance value is close to zero, the variables are multicollinear. As a rule of thumb, a tolerance of less than .20 indicates a problem with multicollinearity. Variance inflation factor (VIF) VIF is the inverse of the tolerance (1/tolerance). VIF has a range 1 to infinity.

37

CHAPTER 4

ANALYSIS AND DISCUSSION

4.1.

Preparation Research.

The approach taken in this research is through the methods of quantitative statistical analysis of the correlation and regression analysis to determine significant factors affecting the accuracy of early cost estimates.

The research data were collected from several sources of relevant literature. By examining the duplication of variables-variables obtained from several sources of literature and examine their relevance to the purpose and limitations of the study, the factors that affect the accuracy of the preliminary cost estimate can be collected as much as 72 variables.

Furthermore, to complement and verify the research variables is done by interviews with experts. Three experts consisting of two practitioners and one academic has reviewed and provided critical inputs in this study. Expert interview process is focused on getting the factors that can not be obtained from literature sources and confirmed based on their experience whether these factors are relevant to the purposes and limitations of research that has been determined. Through expert interviews obtained the factors that affect the accuracy of preliminary cost estimates as many as 54 factors.

38 Here is a list of 54 factors that affect the accuracy of the preliminary cost estimate after an interview with an expert who will then become the basis of questions in the questionnaire:

I 1. 2. 3. 4. 5. II 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

NATURE OF THE PROJECT Type of the project Size of Project / Scope of Project Complexity of project Geographical location / Site location The contract procurement system THE LEVEL OF DATA AND INFORMATION AVAILABLE Completeness of cost information Accuracy and reliability of cost information Applicability of cost information Availability Project information / Documents used in Preparing the estimate Capacities of project Time allowed for Preparing the estimate Technology: Method of construction Project Processes Project design criteria Complexity of design and construction Utility resources and supply conditions Lead time for delivery of materials Project Schedule Project strategy Frequency variations of construction Client's financial situation

39
22. 23. 24. 25. III 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. IV 43. 44. 45. Owner's cost Owner's experience level Engineer / Designer 's experience level The type of structure Techniques and Methods Used Alignment of estimate methodologies, with available project information Is the estimate work process formally define and Followed? Categorized and a formal structure to prepare the estimate Utilization of a checklist to Ensure completeness and technical base Standard procedure is to update the cost data of the project Documentation of information used in Preparing the estimate Involvement of other resources in Preparing the estimates Level of involvement of the project manager Level of team integration and alignment Review and acceptance of estimate by Appropriate parties Attitude / culture forward changes Method used to determine contingency comparison estimate, unit techniques, functional approach, resource estimate factor estimate ABILITY OF ESTIMATOR Relevant experience of the estimating team Quality level of Estimator / Estimator Expertise Attributes of Estimator

40
46 V 47. 48. 49. 50. 51. 52. 53. 54. Acquisition and application of expertise WHILE OTHER FACTOR Considered Preparing the Estimate Impact of governmental requirements / regulation Work force Labour productivity Tax and insurance Money factors (currency, inflation, etc) The prevailing economic climate The nature of the competition / Bidding Competitiveness Logistic for engineering and Constructions

Table 4.1 Factor affecting the accuracy of early cost estimate after interview with expert. (Source : Analysis result expert questionnaire)

Collecting research data in the form of primary data done with the survey method. The design of the questionnaire in the form of checklists factors that affect the accuracy of preliminary cost estimates as the independent variable (variable X) and dependent variable (variable Y) of the accuracy of preliminary cost estimates on projects that have been addressed. Research questionnaire distributed 70 questionnaires to the respondents sebayak research. Respondents who became a target for research data collection focused to the project owner, the contractor and the project consulting Services Company that has been experienced in the preparation of preliminary cost estimates of construction projects.

4.2.

Data Analysis Questionnaires

The questionnaire collected as many as 50 samples. Category respondent is cost estimator, project manager, construction manager, or site manager who have experiences at least 5 (five) years in building construction project in Jakarta. Profile of respondents can be seen in Table 4.1. From the questionnaire, the data collected

41 and then entered into the computer made the tabulation of data in Excel format to facilitate the analysis process. The process is treated with assisted statistical analysis software SPSS ver.17. The first analysis is the correlation analysis to measure the strength of the relationship between variables X with variables Y And then proceed with regression analysis to obtain regres model of the factors that affect the accuracy of preliminary cost estimates.
No. 1. 2. 3. Type of Company The project owner Consultant Contractor Total 3 28 19

Table 4.2. Profile of respondents

4.2.1

Analysis Correlation Correlation analysis of the data obtained is performed to find the strength of

association between two variables, namely the factors of influence and level of accuracy of the preliminary cost estimates. Data compiled by the survey results tabulated in Microsoft Excel and then transpose into SPSS. From the questionnaire data have been obtained, results of correlation analysis can be seen in Table 4.3. as follows:

Variable symbol X1

THE EARLY Factors COST Accuracy of Estimate Type of the project Pearson Correlation Sig. (2-tailed) N 0, 048 0, 742 50 0.279 * 0, 050 50 0, 091 0, 531 50

X2

Size of Project / Scope of Project

Pearson Correlation Sig. (2-tailed) N

X3

Complexity of project

Pearson Correlation Sig. (2-tailed) N

42
X4 Geographical location / Site location Pearson Correlation Sig. (2-tailed) N X5 The contract procurement system Pearson Correlation Sig. (2-tailed) N X6 Completeness of cost information Pearson Correlation Sig. (2-tailed) N X7 Accuracy and reliability of data / information and document Pearson Correlation Sig. (2-tailed) N X8 Applicability of cost information Pearson Correlation Sig. (2-tailed) N X9 Availability project information / Documents used in Preparing the estimate X10 Capacities of project Pearson Correlation Sig. (2-tailed) N Pearson Correlation Sig. (2-tailed) N X11 Time allowed for Preparing the estimate Pearson Correlation Sig. (2-tailed) N X12 Technology: Method of construction Pearson Correlation Sig. (2-tailed) N X13 Project Processes Pearson Correlation Sig. (2-tailed) N X 14 Project design criteria Pearson Correlation Sig. (2-tailed) N X15 Complexity of design and construction Pearson Correlation Sig. (2-tailed) N X16 Utility resources and supply conditions Pearson Correlation Sig. (2-tailed) N 0.769 ** 0, 000 50 0, 086 0, 554 50 0, 182 0, 207 50 0, 105 0, 468 50 0, 118 0, 416 50 0.685 ** 0, 000 50 -0, 115 0, 427 50 0, 137 0, 341 50 0, 071 0, 624 50 0, 051 0, 724 50 -0, 051 0, 723 50 0, 091 0, 531 50 0, 010 0, 945 50

43
X17 Lead time for delivery of materials Pearson Correlation Sig. (2-tailed) N X18 Project Schedule Pearson Correlation Sig. (2-tailed) N X19 Project strategy Pearson Correlation Sig. (2-tailed) N X20 Frequency variations of construction Pearson Correlation Sig. (2-tailed) N X21 Client's financial situation Pearson Correlation Sig. (2-tailed) N X22 Owner's cost Pearson Correlation Sig. (2-tailed) N X23 Owner's experience level Pearson Correlation Sig. (2-tailed) N X24 Engineer / Designer 's experience level Pearson Correlation Sig. (2-tailed) N X25 The type of structure Pearson Correlation Sig. (2-tailed) N X26 Alignment of estimate methodologies, with available project information Pearson Correlation Sig. (2-tailed) N X27 Is the estimate work process formally define and Followed? Pearson Correlation Sig. (2-tailed) N X28 Categorized and a formal structure to prepare the estimate Pearson Correlation Sig. (2-tailed) N X29 Utilization of a checklist to Ensure completeness and technical base Pearson Correlation Sig. (2-tailed) N 0, 054 0, 712 50 -0, 078 0, 589 50 0, 056 0, 698 50 -0, 048 0, 738 50 0, 069 0, 632 50 0, 225 0, 116 50 0, 126 0, 384 50 0, 106 0, 462 50 0, 043 0, 765 50 0, 063 0, 666 50 0, 239 0, 095 50 0, 154 0, 286 50 0, 147 0, 309 50

44
X30 Standard procedure is to update the cost data of the project Pearson Correlation Sig. (2-tailed) N X31 Documentation of information used in Preparing the estimate Pearson Correlation Sig. (2-tailed) N X32 Involvement of other resources in Preparing the estimates Pearson Correlation Sig. (2-tailed) N X33 Level of involvement of the project manager Pearson Correlation Sig. (2-tailed) N X34 Level of team integration and alignment Pearson Correlation Sig. (2-tailed) N X35 Review and acceptance of estimate by Appropriate parties Pearson Correlation Sig. (2-tailed) N X36 Attitude / culture forward changes Pearson Correlation Sig. (2-tailed) N X37 Method used to determine contingency Pearson Correlation Sig. (2-tailed) N X38 comparison estimate, Pearson Correlation Sig. (2-tailed) N X39 unit techniques, Pearson Correlation Sig. (2-tailed) N X40 functional approach, Pearson Correlation Sig. (2-tailed) N X41 resource estimate Pearson Correlation Sig. (2-tailed) N X42 factor estimate Pearson Correlation Sig. (2-tailed) N 0, 108 0, 454 50 0, 114 0, 432 50 0, 040 0, 782 50 0, 097 0, 503 50 0.323 * 0, 022 50 0, 169 0, 240 50 0, 086 0, 553 50 0, 163 0, 259 50 0, 130 0, 367 50 0, 010 0, 944 50 0, 060 0, 680 50 -0, 008 0, 956 50 0, 121 0, 404 50

45
X43 Relevant experience of the estimating team Pearson Correlation Sig. (2-tailed) N X44 Quality level of the estimator / Estimator Expertise Pearson Correlation Sig. (2-tailed) N X45 Attributes of Estimator Pearson Correlation Sig. (2-tailed) N X46 Acquisition and application of expertise Pearson Correlation Sig. (2-tailed) N X47 Impact of governmental requirements / regulation Pearson Correlation Sig. (2-tailed) N X48 Work force Pearson Correlation Sig. (2-tailed) N X49 Labour productivity Pearson Correlation Sig. (2-tailed) N X50 Tax and insurance Pearson Correlation Sig. (2-tailed) N X51 Money factors (currency, inflation, etc) Pearson Correlation Sig. (2-tailed) N X52 The prevailing economic climate Pearson Correlation Sig. (2-tailed) N X53 The nature of the competition / Bidding Competitiveness Pearson Correlation Sig. (2-tailed) N X54 Logistic for engineering and Constructions Pearson Correlation Sig. (2-tailed) N
* Correlation is significant at the 0:05 level (2-tailed). * * Correlation is significant at the 0:01 level (2-tailed).

0.447 ** 0, 001 50 0.437 ** 0, 001 50 0, 264 0, 064 50 0.323 * 0, 022 50 0.749 ** 0, 000 50 0, 117 0, 417 50 0, 034 0, 816 50 0, 025 0, 865 50 0, 001 0, 992 50 -0, 007 0, 962 50 0, 058 0, 687 50 -0, 233 0, 104 50

Table 4.3. Coeficient Correlation and significant levels of variables

46 From the table 4.4. above can be seen which factors have a significant correlation of early cost estimate accuracy indicated by the value significant level <0.05.

The following variables-variables with correlation coefficient and a significant degree of confidence:

Coef. Correlation X2: X4: X9: X34: X43: X44: X46: X47: Size of Project / Scope of Project Geographical location / Site location. Availability project information / Documents used in Preparing the estimate. Level of team integration and alignment. Relevant experience of the estimating team. Quality level of the estimator / Estimator Expertise. Acquisition and application of expertise. Impact of governmental requirements / regulations. 0.279 0.769 0.685 0.323 0.447 0.437 0.323 0.749

Significant Level 0.05 0.000 0.000 0.022 0.001 0.001 0.022 0.000

Table 4.4. Factors that have a significant correlation of early cost estimate accuracy (Source : Result of correlation analysis)

From the results of correlation analysis the correlation coefficient obtained from the independent variable on the relationship with the dependent variable showed a positive number (possitive / direct correlation), that is the change of independent variable will cause a change in the dependent variable on a regular basis and direction. So it can be concluded that if the performance of the variable X above which are all factors that affect the accuracy of the preliminary cost estimate increases, the accuracy of the preliminary cost estimate will increase.

47 4.2.2 Regression Analysis

Regression analysis conducted with the help of SPSS software ver. 17. Through regression analysis can be seen how the strength of the relationship between one or several independent variables with a dependent variable via a regression model produced.

In this study it is assumed that the relationship between variables, independent variables and the dependent variable is the liner, so the regression analysis conducted in this study were linear regression analysis. In this regression analysis, significant variables (X2, X4, X9, X34, X43, X44, X46 and X47) that have been generated through correlation analysis of determinants included as independent variables to generate a regression model. With regard to the value Adjusted R Square and Coefficient Index it will produce a regression model that can express the relationship between the independent variables significantly to the dependent variable.

Results of initial regression analysis, can be seen in the following table:

Table 4.5: Model Summary

Table 4.6: Diagnostic collinearity

48 Table 4.7: Coefficient

On table 4.5 model summary, See figure Adjusted R Square is 0.711, it shows the numbers Adjusted R Square good enough (near to 1).

On Table 4.6 Diagnostic collinearity, See figure Condition Index The biggest is 37,213. Terms of model validation (test multicollinearity), Condition Index must be 18. Therefore there are four variables that must be removed in order to qualify validation. How to remove the variable is to look at the value Sig on Table 4.7 coefficient. Remove the 4 (four) variables have the high value Sig X46: X34: X47, and X43. Then repeated the regression with independent variables X4, X44, X9, and X2.

The results of the subsequent regression analysis, can be seen in the following table:

Table 4.8 : Model Summary 1

49 Table 4.9 : Diagnostic collinearity 1

Table 4.10: Coefficients 1

On Table 4.8. Model Summary 1, See figure Adjusted R Square is 0.713, this suggests there is an increasing number Adjusted R Square become better. On Table 4.9. Collinearity Diagnostic 1, Looks Condition Index produced the largest is 17.013, which it has been qualified test models so that we can make it to see the regression model coefficient in column B in Table 4.10. Coefficients 1.

The resulting regression model is as follows: Y = 0.326 + 0.355 X4 + 0.243 X9 + 0.174 X44+ 0.054 X2

In the research, the data collected are often biased data (outliers) so it needs to be re-examined data dissemination through Scatter plot - Regression Standardized Predicted Value. The data is the outlier can be seen from the position data that are beyond the threshold regression line with Confidence interval 95%. Therefore the

50 initial data (raw data) Should be re-examined and the data that acts as outlier should be removed and then re-regression analysis was repeated so that the final regression model can be generated.

The process of removal of outlier data and by performing repeated regression analysis will increase the value adjusted R Square so that means that the reliability of the resulting regression model will be better. On Charts - Y1 by * zpred scatterplot in the regression analysis output can be seen early regression scatter diagram as shown in figure 4.1 follows:

Figure 4.1. Scatter plot - Regression Standardized Predicted Value

In Figure 4.1 shows that the respondent no. 29 (R29) are beyond the threshold regression, meaning that this data should be released. After the data no. 29 issued, performed the regression again to get a better regression model. The results of the subsequent regression analysis can be seen in the following table:

51 Table 4.11 : Model Summary 2

Table 4.12 : Table Diagnostic collinearity 2

Table 4.13 : Table Coefficients 2

On Table 4.11. Model Summary 2, See figure Adjusted R Square is 0.791, meaning there is a significant improvement from the previous model by removing data no. 29.

52 The resulting regression model is as follows: Y = 0.099 + 0.382 X4 + 0.245 X9 + 0.194 X44 + 0.056 X2

The process of removing data outliers and regression analysis is repeated until all variables were obtained within the limits of regression. Outlier data that should be eliminated in addition to data no. 29 is a data No.30, no.6, no.25, no. 24 and no. 40, so the final regression analysis performed with data as much as 44 data. Final data obtained from the process which will then be used in the regression model. Here are the final results of regression analysis produced:

Table 4.14: Model Summary (End)

Table 4.15: Coefficients (End)

53

Figure 4.2. Scatter plot - Regression Standardized Predicted Value (End)

On Table 4.14. Model Summary (end), Can be seen the value of Adjusted R Square into 0.854, this shows the increased value of the Adjusted R Square initial (0711). On Figure 4.2, the spread variable in the scatter plot shows the variables inside the threshold of the final regression model regression so that we can produce are as follows:

Y = 0.108 + 0.411 X4 + 0.208 X44 + 0.156 X9 + 0.096 X2

Where: Y : Accuracy of early cost estimate. X4 : Geographical location / Site location. : Availability project information / Documents used in Preparing X9 the estimate. X44 : Quality level of the estimator / Estimator Expertise. X2 : Size of Project / Scope of the project.

54 4.2.3 Model Test

4.2.3.1 Linearity Test ( F Test )

F Test or Analysis of Variance (ANOVA) Aimed to test whether the resulting regression model is linear or not. F test is done by looking at the data in table ANOVA below that obtained from the regression analysis output.

Table 4.16 : ANOVA

Hypothesis: H 0 = not a linear regression model H 1 = linear regression model Based on the output ANOVA obtained the Sig = 0.000 = 0% less than (5%) or the value of F = 63.762 more than the F table = F (4; 39; 0.05) = 2.61, then H 0 rejected and H 1 accepted. Thus is a linear regression model and the resulting regression model is highly significant.

55 4.2.3.2 t Test

t-test (Student-t Distribution) Aims to determine the trust level of each independent variable in regression equations or models used in predicting the value of Y. usually t test is also called coeffisient test.

Hypothesis: H 0 : regression coefficients not significant H 1 : Significant regression coefficients

On Table 4.15. coefficient (end), can be seen the value of Sig. range is 0.000 - 0.047 or smaller than 0.05, and the value of t ranged between 2.052-8.538 more than value t table = t (39; 0.05) = 1.686. then H 0 rejected or H 1 accepted. So the regression coefficient for independent variable is significant

4.2.3.3 Autocorrelation Test (Durbin Watson Test)

Auto correlation test done to measure whether there is autocorrelation between variables in different samples. Autocorrelation test with thresholds Durbin Watson -2 To +2 to determine the presence or absence of residual correlation or auto correlation of the resulting regression model. Value Durbin Watson generated through the output of regression analysis can be seen in table. 4.14. model summary (end), namely 1.891. The values are still within the limits of the value terms Durbin Watson ie on the interval -2 < Durbin Watson <+2, So it can be concluded that the resulting regression model does not occur auto correlation.

56 4.2.3.4 Multicollinearity Test

Based on the output Coefficients earned value tolerance in the between 0.646 0.901 is less than 1 and value of VIF in the between 1.109 - 1.549 is less than 10. It means there is no multicollinearity.

4.3

Result of Research

Based on the results of research conducted, some variable determine accuracy of early cost estimate (Geographical location / Site location, Availability project information / Documents used in Preparing the estimate, Quality level of the estimator / estimator Expertise and Size of Project / Scope of project) on the project construction is valid, to see the results of model testing on the F test, Q test, autocorelation test and multicollinearity tests.

57

CHAPTER 5

CONCLUSION AND RECOMMENDATION

5.1

Introduction

Early cost estimate are cost estimates prepared in the early stages of planning the project before the completion of engineering design made, therefore early cost estimates have an important function as a starting point and a benchmark for project planning and controlling in the future project stages. Quality early cost estimate can be measured from its level of accuracy. Early cost estimates are prepared based on limited information and supporting data, it causes in early cost estimates has the lowest level of accuracy than any other types of cost estimates.

In this research, researcher want to identifying factors that affect the accuracy of early cost estimates of construction projects and identifying relationship between the factors and the accuracy of early cost estimates.

5.2

Conclusion

From the literature review, researcher found 72 factor affecting the accuracy of early cost estimate. These factors reduce to 54 factors after interview with expert,. Expert interview process is focused on getting the factors that cannot be obtained from literature sources and confirmed based on their experience whether these factors are

58 relevant to the purposes and limitations of research that has been determined. The 54 factors that has been determined will be a basic question in the questionnaire that will be spread to the respondent that would later analyzed using correlation analysis and regression analysis.

From the results of the analysis in chapter 4, obtained 4 (four) significant factors that affecting the accuracy of early cost estimate. These factors based on level of affecting on accuracy, the sequence is as follows: 1. Geographical location / Site location. 2. Quality level of the estimator / Estimator Expertise. 3. Availability project information / Documents used in Preparing the Estimate. 4. Size of Project / Scope of the project

The relationship between the variables determining the accuracy early cost estimate can be seen in the regression model below:

Y = 0.108 + 0.411 X4 + 0.208 X44 + 0.156 X9 + 0.096 X2

The other variables considered less influence on early cost estimate accuracy, but it does not mean no effect at all. Cost estimator should always take into account all the variables that affect accuracy in making early cost estimate, with a focus on the factors the most influential factors as mentioned above.

59 5.3 Limitation Of Research

Result of this research valid when used to determine early cost estimate at building-construction project or similar project in Jakarta, Indonesia.

5.4

Recommendation For Future Research

Significant factors resulting from this research can be used as reference for cost estimator when they prepare an early cost estimate for the next project and that results can be developed in subsequent research that is for different types of projects or for other types of cost estimates i.e. analysis of factor affecting the accuracy of early cost estimate in highway construction project.

59

REFERENCES

AACE International Recommended Practice No. 17R-97. (2009), Cost Estimate Classification System. AACE Inc., AACE International Recommended Practice No. 34R-05. (2008), Basis of Estimate. AACE Inc., AACE International Recommended Practice No. 40R-08. (2008), Contingency Estimating General Principles. AACE Inc., Aibinu, A. A., and Pasco, T., (2008), The accuracy of pre-tender building cost estimates in Australia, Construction Management and Economics, Melbourne, Australia. Ashworth, A and Skitmore, R. M. (1982), Accuracy in Estimating, Occasional paper no. 27, chartered Institute of Building, London Asiyanto. (2003), Construction Project Cost Management. PT. Pradnya Paramita. Jakarta,. David Baccarini. (2004), Accuracy in Estimating Project Cost Construction Contingency A Statistical Analysis. RICS Construction and Building Research Conference. RICS. London, Iman Soeharto. (1999.), Manajemen Proyek (Dari Konseptual Sampai Operasional), Jilid 1. Edisi Kedua, Erlangga, Jakarta, Iman Soeharto. (2001), Manajemen Proyek (Dari Konseptual Sampai Operasional), Jilid 2. Edisi Kedua, Erlangga, Jakarta,.

60 Martin Brook. (2004), Estimating and Tendering for Construction Work. Third Edition, Elsevier, Amsterdam,. Morris, M. R. (1990). Improving the accuracy of early cost-estimates for federal construction projects, Building Research Board, National Research Council, Washington, D.C. Oberlender, G. D., et al. (1998a). Improving early estimates. Research Summary 1311, Construction Industry Institute, Austin, Texas. Oberlender, G. D., et al. (1998b). Improving early estimatesBest practices guide. Implementation Resource 131-2, Construction Industry Institute, Austin, Texas. Oberlender, G. D., et al. (1998c). Improving early estimates source document. Research Rep. 131-11, Construction Industry Institute, Austin, Texas. Oberlander, G. D., and Trost, S. M. (2001). Predicting accuracy of early cost estimates based on estimate quality. J. Constr. Eng. Manage. Oberlender. G. D., (2000), Project Management for Engineering and Construction. 2nd Edition. McGraw-Hill Book Company Inc. New York, O. Babalola and O.Aladegbaiye. (2006), Determining Appropriate Sum for Building Projects. RICS Construction and Building Research Conference. RICS. London, Phillip F. Ostwald. (2001), Construction Cost Analysis and Estimating. Prentice Hall, New Jersey,. Richard Fellows and Anita Liu. (2003), Research Methods for Construction. Second Edition, Blackwell Science Ltd, Oxford, UK,. Robert L. Peurifoy and Garold D. Oberlender. (2004), Estimating Construction Cost. Fifth Edition, McGraw-Hill, Boston.

61 Roger Flanagan and George Norman. (1993), Risk Management and Construction. Blackwell Science Ltd. London UK,. Skitmore, Martin and Stradling, Steve and Tuohy, Alan and Mkwezalamba, Harry (1990) The accuracy of construction price forecasts. Technical Report, Department of Surveying, University of Salford. Skitmore, R. M and Picken, D (2000), The Accuracy of Pre Tender Building Price Forecast: An Analysis of USA Data, Australian Institute of Quantity Surveyors refereed Journal 4 Stephen Ogunlana, Antony Thorpe (1991), The Nature of Estimating Accuracy: Developing Correct Associations, Great Britain. Steven M Trost and Garold D. Oberlander (2003), Predicting Accuracy of Early Estimate Using Factor Analysis and Multivariate Regression. Journal of Construction Engineering and Management. ASCE, 2003. Tarek Hegazy. (2003), Computer-Based Construction in Project Management. Prentice Hall, New Jersey. Yean Y. L., and Jina, H. S. B., (2001), Improving The Accuracy Estimates of Building of Approximate Projects, Building Research and Information, Singapore

APPENDIX A
Expert Interview

Research Questionnaire

INTRODUCTION Early cost estimates are cost estimates prepared in the early stages of the project is at feasibility study stage until just before the detailed design completed. Early cost estimates based on data and information is still limited as the initial concept of an outline of the scope of the project work. The existence of the limited available data and information is one of the preliminary cost estimate has lower accuracy compared to other types of cost estimates. Early cost estimate for the cost estimates made in the first phase of the project cycle, so that early cost estimates are a starting point and a benchmark in determining the steps to be done next, for example in terms of planning and strategy development, determining project funding strategies, filter out projects which are the potential of its business, determines strategy project will be implemented, as a reference for the development of cost estimates and the subsequent engineering design, etc.. In the process of preparing early cost estimate, views of the importance of early estimates as a basis for decision making on the one hand, and on the other side where the accuracy of early cost estimates lower than other types of cost estimates, creating a stark contrast to conditions that are often experienced in the early stages of development conceptual project. That condition underlying author conduct this research, Thus the title of research is : ANALYSIS OF FACTORS AFFECTING THE ACCURACY OF EARLY COST ESTIMATE IN CONSTRUCTION PROJECTS The quality of cost estimates associated with the accuracy and completeness of its elements. The accuracy of the estimated project cost depend on: nature of project, the level of data and information available, techniques and methods used, ability of the estimator, and other factor considered while preparing the estimate. Relationship accuracy of the supporting elements can be described as follows.

Y
Accuracy

Influence Factors X

This study aimed to identify factors that influence the accuracy of preliminary cost estimates for construction project will then be analyzed by quantitative methods in order to find the significant factors on the accuracy of preliminary cost estimates. These factors result from the analysis are expected to provide input on matters that need to be prepared, techniques and methods used, the required procedures and other matters related to the process of preliminary cost estimates in order to produce a preliminary cost estimate is accurate.

Sincerely, Researcher: Dodi Ardianto, ST email: dodiardianto@hotmail.com Supervisor: Assoc. Prof. DR. Razali Abdul Hamid

DATA OF EXPERT

1. Name

2. Compa ny

3. Profession / Department

4. Work Experience

Year

5. Project Name

6. Project Locat ion

7. Year Project Completed

COMMENTS AND SUGGESTIONS

Jakarta,

/ 2011

( ..............................)

Questionnaire Instructions 1. 2. Filling this questionnaire based on your experiences and perceptions. Filling the questionnaire was done by answering the question: "Do You Agree Variables Below Affecting Accuracy of Early Cost Estimate?" 3. 4. 5. Optional answers were: YES or NO Filling an answer by providing the sign of [] In the column provided. When there are other things that need to be added, please enter the last row in the table

"Do You Agree Variables Below Affecting Accuracy of Early Cost Estimate?"
I NATURE OF THE PROJECT Yes 1. 2. 3. 4. 5. 6. 7. Type of the project Size of project / Scope of project Complexity of project Other project characteristics Geographical location / Site location The contract procurement system The prevailing economic climate No

Other factors that need to be added

II

THE LEVEL OF DATA AND INFORMATION AVAILABLE Yes No

8. 9. 10. 11. 12.

Completeness of cost information Quality of information Accuracy and reliability of cost information Applicability of cost information Availability Project information/Documents used in preparing the estimate

13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

Capacities of project Time allowed for preparing the estimate Technology; Method of construction Processes project Piping & Instrumentation Diagrams Amount of special work likely production time Project design criteria Complexity of design and construction Utility resources and supply conditions Heat and material balances Mechanical equipment list Lead time for delivery of materials Project Schedule The form of procurement and contractual agreement Project strategy The extent of completion of pre-contract decision Frequency of construction variations Clients financial situation Owners cost Purpose and intended use of estimate Owners experience level Engineer/Designers experience level Environmental assessment and workforce The type of structure

Other factors that need to be added

III

Techniques and Methods Used Yes No

37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56

Alignment of estimate methodology Is the estimate work process formally define and followed? Formal structure to categorized and prepare the estimate Utilization of checklist to ensure completeness and technical basis Standard procedure to update cost data of project Documentation of information used in preparing the estimate Involvement of other resources in preparing the estimates Level of involvement of the project manager Level of team integration and alignment Review and acceptance of estimate by appropriate parties Attitude/culture forward changes Method used to determine contingency the conference estimate, comparison estimate, graphical relationships, unit techniques, functional approach, resource estimate factor estimate exponent estimate

Other factors that need to be added

IV

ABILITY OF ESTIMATOR Yes No

57. 58. 59. 60. 61

Estimator expertise Relevant experience of the estimating team Quality level of Estimator Attributes of Estimator Acquisition and application of expertise

Other factors that need to be added

OTHER FACTOR CONSIDERED WHILE PREPARING THE ESTIMATE Yes No

62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72.

Impact of project type Impact of contract type Impact of project schedule Impact of governmental requirements / regulation Work force Labour productivity Tax and insurance Money factors Bidding climate The nature of the competition / Bidding Competitiveness Logistic for engineering and constructions

Other factors that need to be added

APPENDIX B
Questionnaire

Research Questionnaire

INTRODUCTION Early cost estimates are cost estimates prepared in the early stages of the project is at feasibility study stage until just before the detailed design completed. Early cost estimates based on data and information is still limited as the initial concept of an outline of the scope of the project work. The existence of the limited available data and information is one of the preliminary cost estimate has lower accuracy compared to other types of cost estimates. Early cost estimate for the cost estimates made in the first phase of the project cycle, so that early cost estimates are a starting point and a benchmark in determining the steps to be done next, for example in terms of planning and strategy development, determining project funding strategies, filter out projects which are the potential of its business, determines strategy project will be implemented, as a reference for the development of cost estimates and the subsequent engineering design, etc.. In the process of preparing early cost estimate, views of the importance of early estimates as a basis for decision making on the one hand, and on the other side where the accuracy of early cost estimates lower than other types of cost estimates, creating a stark contrast to conditions that are often experienced in the early stages of development conceptual project. That condition underlying author conduct this research, Thus the title of research is : ANALYSIS OF FACTORS AFFECTING THE ACCURACY OF EARLY COST ESTIMATE IN CONSTRUCTION PROJECTS The quality of cost estimates associated with the accuracy and completeness of its elements. The accuracy of the estimated project cost depend on: nature of project, the level of data and information available, techniques and methods used, ability of the estimator, and other factor considered while preparing the estimate. Relationship accuracy of the supporting elements can be described as follows.

Y
Accuracy

Influence Factors X

This study aimed to identify factors that influence the accuracy of preliminary cost estimates for construction project will then be analyzed by quantitative methods in order to find the significant factors on the accuracy of preliminary cost estimates. These factors result from the analysis are expected to provide input on matters that need to be prepared, techniques and methods used, the required procedures and other matters related to the process of preliminary cost estimates in order to produce a preliminary cost estimate is accurate.

Sincerely, Researcher: Dodi Ardianto, ST email: dodiardianto@hotmail.com Supervisor: Assoc. Prof. DR. Razali Abdul Hamid

RESPONDENT DATA & REFERENCE PROJECT

1. Respondent Name

2. Compa ny Na me

3. Profession / Department

4. Work Experience

Years

5. Project Name

6. Project Locat ion

7. Year Project Completed

COMMENTS AND SUGGESTIONS

Jakarta,

/ 2011

( ..............................)

Questionnaire Research In this questionnaire is expected to be completed by respondent under conditions where the project is on early stages conceptual project before detail design produced, by answering t h e q u e s t i o n : how the factors below on your project? Check [ X] Or CLICK one time in the column provided on your answer to one of the options. Please read carefully and thoroughly before giving answers.
I NATURE OF THE PROJECT

What Was Known About The Project? Answer (1) ( 2) ( 3) ( 4) ( 5) : : : : : Do not Know L i t t l e Bi t K n ow A v era g e K no w Ma ny K no w R ea l ly K n ow 1 2 3 4 5

1. 2. 3. 4. 5. II

Type of the project Size of project /Scope of Project Complexity of project Geographical location / Site location The contract procurement system THE LEVEL OF DATA AND INFORMATION AVAILABLE

How is the availability of data and information below on your project? Answer (1) ( 2) ( 3) ( 4) ( 5) : : : : : Not Available At All Fe w A va i la bl e Fa i rl y Av a il a bl e Ma ny Ava ila b l e Hi g hl y Av a i la bl e 1 2 3 4 5

6. 7. 8. 9. 10.

Completeness of cost information Accuracy and reliability of cost information Applicability of cost information Availability Project information/Documents used in preparing the estimate Capacities of project

11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. III

Time allowed for preparing the estimate Technology; Method of construction Processes project Project design criteria Complexity of design and construction Utility resources and supply conditions Lead time for delivery of materials Project Schedule Project strategy Frequency of construction variations Clients financial situation Owners cost Owners experience level Engineer/Designers experience level The type of structure Techniques and Methods Used

How the use of the techniques and methods of early cost estimates on your project? Answer (1) ( 2) ( 3) ( 4) ( 5) : : : : : Not Used At All S l i gh t ly Us e d Us e d Of t e n Us ed A lwa ys Us ed 1 2 3 4 5

26. 27. 28. 29. 30. 31.

Alignment of estimate methodology with available project information Is the estimate work process formally define and followed? Formal structure to categorized and prepare the estimate Utilization of checklist to ensure completeness and technical basis Standard procedure to update cost data of project Documentation of information used in preparing the estimate

32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. IV

Involvement of other resources in preparing the estimates Level of involvement of the project manager Level of team integration and alignment Review and acceptance of estimate by appropriate parties Attitude/culture forward changes Method used to determine contingency comparison estimate, unit techniques, functional approach, resource estimate factor estimate ABILITY OF ESTIMATOR

How the skills and experience of estimator in your project? Answer (1) ( 2) ( 3) ( 4) (5) : : : : : Very Bad Bad A v era g e Good V e ry G o o d 1 2 3 4 5

43. 44. 45. 46

Relevant experience of the estimating team Quality level of Estimator / Estimator Expertise Attributes of Estimator Acquisition and application of expertise

OTHER FACTOR CONSIDERED WHILE PREPARING THE ESTIMATE

How the factors below related to the preparation of cost estimates on your project? Answer (1) : Does not affect ( 2) : L i t t l e b i t a f f ec t ( 3) : Af f e c t ( 4) : Mu c h A ff ec t ( 5) : V e ry Af f e c t 47. 48. 49. 50. 51. 52. 53. 54. Impact of governmental requirements / regulation Work force Labour productivity Tax and insurance Money factors (currency, inflation, etc)

The prevailing economic climate The nature of the competition / Bidding Competitiveness Logistic for engineering and constructions

Questionnaire part II:

VI

How big is the accuracy of preliminary cost estimates generated in your project?

(Put a [X] Or CLICK one time in the column provided on one of your answer choices)

Charging indicators: 1. Criteria the accuracy of preliminary cost estimates:

Scale
1 2 3 4 5

Quantitative Assessment
- 50% Y - 30% or + 50% Y + 100% - 20% Y - 10 % or + 20% Y + 50% -10 % Y - 5 % or +10% Y + 20% -5 % Y - 3 % or +5 % Y + 10% -3 % Y + 5%

Level Accuracy
Inaccurate Less accurate Fairly accurate Accurate Very Accurate

2. Calculating the accuracy of preliminary cost estimates can be measured by the formula: Br - Be Y = ----------------- x 100% Be Where: Y = The accuracy of the preliminary project cost estimate Br = Actual Cost Be = Cost Estimates

APPENDIX C
F Table and t Table

APPENDIX D
Data Questionnaire

t
h

+ t

*
d

tf

o rf

rf

t
tl. d

rt
f

t
6

t
q q

+
t
N lf

*
o

*
!t

t a tt
tl 6 !t lf D

tf

it

rf

t
lf

tf

t
b

$
!t q

e rf
6 d 6 tl h 6 N N

rt

t
!t

*
!r
!t q

+ t
t
q

t rt t

tl' \t
tl-

t o o

+
N

*
6

tf

tf f

6 N

t t t

+ c n
e
N

n
t

*
!f

t
b

rt t
6

rf o o

t o t

f d

rf
f

*
!f tf N N q

t rf
g

*
t
h 6 4

{'
tf

a
!i

n
t o o

rl

s o

!.

*
t o t t

\t

n
rf
!l

tt

{
*
rf t
6

lf

*
t o

c
lf 6 6

*
n
C

t
!t tf

t tt
g

tf

o
!
d

*
n

!r
o
h

o o o o
qt

o n t !r
rf
!t h

t
d N

t t

*
d

*
6 I

t
h 6 N

n o
rt'

\t o

tr n

D N tl

rf o

rf
6 N !f
h

6 D 6 q

+ rt
tf

+
tl C

o o

+
rf rl
f

rf

N f if

c/
6t

n
t

!f !t

!f

tl !f

l0.

rf

<l

tl

+ rt
t t
q 6 b q

o
t
6

al

s t
t

+
o
4

rt
6 h 6 ll

t
d N

tf tf

6 !t tl d

t t

tl q

tt
q

fl

rf

ri
d

r?

t
d

t t

t
N lf lf

t
!f

rt

rl
lf

t
tl

rt
!t !f d

{
a rf

o
q

v
rf

rr

rr
rf

s
6

t
il 6

{5

x e x 3 x x o x 6 x x !x

O I

x x x x x x x x x x x x x x x x x x x

c
E B

o c E

I o

I
e
ll o
E

o g c q

p
E

a g q
5
c

E f
&
a L a c

s
g
E

E E o
E
E
E.

E c

5 e
E

5 H
c

p
B *. 6
E
nt

E ,c

s o
E o
a

E
E

c a E

E E

a o 5
o E

2
G

$
o c

E 5
o

E o
6

la
-!!

I
-q

E
.9

{
5 e ti
t b c o e

E c

E
8.

b 5 E E o b 5
E .q

I !
E p

g
d

g E c

g 5

g
E

&
.s

g
E

.t

{
,E

fl 2

a o U s t E a ? s c b t E' o a E E g o a o .gr s s ; b g fi fi E I R a v g E $ v ii E * v 3 s s -a g g g 6- 6 I 6 I 3 E T I E' 9 I 'a E s

tt o

; I

g *. B

a a

E
a

o J E

fl

E
J

b
.3

fi
c

B
R 6 t

tE
! c -!!

: : E I I o E B I tc { : I a c
p
J

6 E 8. 5

a e
g
E

E
6

b
e

t: a 3 g g J
e

b $ I
5 H i6 F
t3

= E F .t

5 g

rl t

a
t
h

tf

t t
6

*
6

tf

rl

$
6

o
b !f
r?

6 6 6

rf ?
q q 6

N q

!r +

t
f

o o

6 6 tl 6 !t tt tf h

rt

t a
lf b

t
6

*
t t

o
f

lf !f

t t t
q

t
q

t
N

0
m
h

n
t
h

t
q

t t
tf

{
I f

t c
q

? t
6

*
!t

rf

tl

tr t
rt
I N 6 6 \l q
h

t
f

rt
rf

t
6

tt
f

+ rf rl
o
tt !t ! f

*
f

* *
q !t tf

t
!t

s
I

!t

s rr
6

b 6 6

d 6

!l

tf

+ rf
rf
!f

+ +

o
6

n
o

!t ?

\t

s rf
t

t
6

*
n
b

n
tf

t
tf

n
lf 6

tt

t t

+
N

c
d

n c
!t

!t

*
D rl
h

*
rt
h

*
t
tt

!r
t

o
6 ro 6 6

*
h

rf

rl
6

t
6

n
t o

!'.

n
rf tf

*
rl tt
q

tt
tt

lf

o
qt

rf rr t o
C

lr r|
rt

t t

rt

tr
6

o
o 3
cl

*
rl o
6 6

*
rf t

t
q

!t

rf t

rf t

+ + t
o
I

? rl
q

+ t
N

+ t
o t
g

rt
f
b

t tt !t

t o

*
n +
q q tl h q

!f

rf
N

q q

t rf
6

s
6

* *
t
!f

tf

!F

rf o

s
t
!f q

t o

$ t
q

N q

+ t

rt

*
a o x 3 o x

* *
6

rl

x o x 6 x 6 x

o 6 x

I I

o I

o i a x x

N ! I

o I

t o 5 5 6 t 5 a I o I O ? x x x x x x b x x x x
I I

ta

a
E C
E

A
a

g
E

E
o

E o

c a $ 8. a c a L e n
g
c

d o

s o
c

a a g
E o o E

E ! a

i
l, o
E o
G

I c
3.
F

! E

E E e s t e g a g

I
5

E t

8
d

s a 6

Ic

c
E

E
8
h.

b
_o

b
6 6

I t u f ts I
E
.E

E B E p B
c
E

o c a

g e
I

8. U

e a
8.

E I

u
=

g
E

b o
8.

s s u

E E i I E
g
J

e o

c
fr o
e

t tc !
E

-g c

E
&

b
n

g
.g g

I g E
h
E
E
!

b a
E

E I

c E

.c

5 o

x { E

5 E 5
c

I I a a
9 g

_s

b
E
6 o g

s
s
d o c o

s
E n
E

s c

E
5

a h

; .
L

E
8. 0

a
t

& B

E E

E 6 5

6
P 5

I l E
u o
E

fi 3

- g

E
6

b g

E
.8

E E

9 I

I ; c
t

c t

5
B

=
t!t

I 6
c o

t
f

I o

a T F

!t tf

ir o

t it
6 !l !t tl

c
N

tf

tl.

f lf ll E

*
o
6 6

n
a t
6

n c n t
t
b

o o
tt
d

{'
m t
h

n
rt
q

rf t

tt
6

t t
d

tf

o
6 h !f

!t

t o

t
6

n rt
q

o
tl

rr
d

n t
!l N

tr o
6

!r o o
N

*
if

o o
o

!t

+ n !r t
rt
if

rf

a t
b

t
b

!a

rl

+
t !t
!f

lr
o
N h

*
!f

o
tf

n + t
f

t o o

rt
h

tr o
rf $ rt
!l

rf t fr rt
f

!r rl
rt
q

o
t

tf

+ n
t
h

*
tf

tt

t rf t
!a

rf

!r
q q

f !

t
t3

$ $ q f

*
t 4

*
tf b 6

!r n
6 N

rf

ll

*
t

$ tl q

+ t

rt ? t
f

o ?

!f

lf

rt rl

n a o
h

s
tf

n + t
6

rt
6
d

tf

t o rf c tt rl
h h

*
t rt
tl b tf

t
!f

rt

*
c o
!t
if

n s s
N
h

*
n

!f

+ + rf t
rf ? t o rt
d

*
o
q 6 6

6 6

n o t
h

rf
6

o
d 6 6 6 6 q $ if

+
!t

*
t

rl t

n rt +

+ v a
b b

o 6 o
cv
6t

n n
6 6 tt

t
h

t
d

It
f

tf !t

rf t

+ t n
t
!f

+
6

*
rl
6 tf 6 !f

n
t
b

GI

c
d

tf

*
6

t
h

t rf
E

* *
h

o
6

rt

n
d

it

*
t t t t t
N b f

lf

+
rt
6
h

{-

rt
a t

t
b

n
N

!}

!t N N

o
b N

6 6 N

t
D !t' 6 6

n t
rf
!f N h 6

6 h

t
4

n t
h

6 !f

x R x

a x

o o e 5 a o O o d t I - I G x 3 x O x x x x x x x x x x x fl x x O x x x x x x x x x
h

s
5
G

o C

E E

s.
g.
.C

p
g
.c E o 5

n 3 o
-6
.E

e s o c

t o

g
E

a c
s 3

a
b
ts

c
5

g 6

c o 6

e
o

s 5 6 a a c I E b t E b e

I {

s
E
ts

; o I
g
E
e

f;
B a
e

t
E

.e

6 E
E E
E
l2

6 e a I
o

E
6

>

o 8.

ts 6

E
E

t
o

t
d

ti g
6 B p I I

E e C
.E

6 g E E
lJ

n
o E

c
b

.1

t
e

D
E

e b E t
B

b a g
o

E t E

I
c .g tI

e E t t 5 P o p

E 3 E

a E
E

e a
o
E

e 6

B '; s

b,

b
.q

fi

& A

It 5

e
e E al
3

!r 5
B
.ll

It
E

I
ts

E
E

3 E P

t p

g
E

F
.g

c g P I
,e

n I !! d 6 I
o

8. a

E
g

.rt

I
6 c

b L E b
F

s e a I B c tt o c
e
E

s
E

I
I

,l

t
.o

*
h

rt
h
h

+
+
q

o o
!f

6 6 6 6

tf

lf 6

rl
6

rr.

m
d

t t
f

t
!t

c ? o

o
6 f 6

s
3

a rl

6 tf

t
tf

6 6

t
ll

6 N

*
!r
o

o
!t

t
h

n
rt
!t

rt
b tf !t

!t !f

t
!t t rf

o rl

t
h

!f

t $
6

*
!l
rF

t
6

tt

? t

*
t

n
h 6 tf

n
h

6 6

n $
b 6

n
o
6

$ rt rt t a
6

tr rt
!

o
!f !

!f 6

t t

rr

rc
g

!?

h il

t
q

*
t
6

+
t
t|

s +
t

t rf rt
6

t
q

s t

t
o

l'

!r tt
h

+ ? rt t
6 6
h

s
I

*
rt

* * *
@ h 6

n
t rt
ra
h

*
tt
ti

6 6 h q

6 6 6

n n
b

ll

n
rl

6 f !l

a t

rf e
6

+
h h

b 4

rr rt t
t
?

+ tt rf rf
i

rt c t

t
!t

tt t

t t t t o

t
6 6

*
n
o

lr

+
6

!t

rl
h

n n n c
o

rt

? rf

rt t

*
rt

o o o

n
@

rt o
d

o o

o
6l

a
N !l h if

o
6 F

o o o t rf
tf

o t
f !f

o sl
o
sl

o 6 o

rf

*
tr
!t

o t
lf

*
+
t

*
rf
tF

+ e

rt
N

$
6 !t lf 6 b

*
t
h

rl t t t

t
o

* *
h

t
o

rl

+ n
6

r3

rf

rf
tt

rt rf rt
6 tf D

+
tt
6

{
rt t a
b 6

*
+ rf

a rt t
b

n rr

tf

a
D t' b

+ t +
h

*
t
h

a
h

tt o
6

*
6 b

tt

tf

!l tl

*
a

6 6

6 6

+ + t

rt

n o
t
I

f
h

o
!t

o
h

*
rf t

t
!f

!'

lts

o o x

d x o x

6 I

t x 0 x 6 x a x x
t 3

o a G a o o E e I D - 5 x Ix x x x Ix I x x a x ! x s x x x x x D

I
.8

i c E s a

E
T

g 5
b. ts

3 o

6
a

g
E.

f, p
o
.E

B o o

A
l9

e
E

s
6
ts

$ E b 5 c a E o P I t c g g .c g E E E c s c o E g J B o E il 5 t a P t a ? b c 3 .c g t c It e b e E F c I g o a 8 o c I o t

! a
g
t.

8.
8.

e E a
6
o

r
s c

I
c

E I

3.

u F

s s a J

z E
E

E E 6 u

c 9 5
E

g
E
E

t
E
o

c
E

e E
t

.g

9 c

6 c o =

t t

b a
E

3 5 E b J
3 3

s c

1'

o o o E E

*. E
F

5 c

E
o P t

a u fi u & b b
5

E b s
E c

d
E E

E
I
E

E
c

c
f

b U

c .c

E
;c

5 E

I o g o 5 E d p 3 g b o a d s e o .g t o e s a d E B E E g o c IJ h. fi u 5 a * .8 I o t = o il p o !F I E E E

e a

*.
E

t -

g e 5

fi

APPENDIX E
Summary Respondent

F rr7
A
.E'

Iri
M
rL
6l
{) .54

cl
G
GI (D

a o
cl

&

J4

J. ll) (n h ()
(l)

o
bl)

Fr

|l)

(t
(D (D

o t* J Gi o

o
d

GI

tr

o'
F{

o
o
tr

(, & f A
E GI 0 o gD
ql

u0 (D

E (D
I

s 6l
ll) bo
GI (A

'tr

E
(l)

& pr

o c 6l

o o
q)

'.o
dE
6

o
<A

()

a qt
-0

o n A M U (\
a tri
cl \o

,rl

rll (D o tr (tk Gl z ll) ad= () a o c) (l) cttr e ('to b0 d '; J GIF (D .o cl ll) c o 9c n () () M gh e E (D E.E (, (D GI ot (t) c)-t o aF o (, tr orS a GI c.v M ql c ('l O o Els \o tr .eE cl c g
|t)

cl

c.l
.h

o
A & A

o d .o cl (D Fl

cl

J4

Cd

J' (D
vh

Gt

b! ll) ()
ll)

6i o tr rt o cl

o
o
{D

F '

o
I
GI GI

GI

GI

cl a cl

|l) 'a A ll)

o 'd
(D

o (D

o at)
tr d
6l
GI

v
cl

o
tr

B
cl

*
ts

art

b0

al
ad

cl Ag F o F4 n) L. l&
GI

o ah

o (D o E trt a 3 o o C) E o d .tt !d a o GI

g
&
k
c(l

o)

*E !l) 11)v o o AM F
ql 6l

z o0

o
c)
U)

SE

o rl o

o0
G!

v,
ch

s E
c.l

.o

C)tr ?EA

o,
0,
c^

a.

z L)d
o
rr.)

g8 o

rl)

ql

c) o z o v
lD

ql

o
al
an

6!

tr ll, L)

o c)
ral

t{ ti

A o{ :/ ot U a lo A (I)d F o ()ri

g
&.<
bo

& A
ct q,
b0

ll.}

(l)

4)

o 6

tri (l
b0
GI

o
d

q)
cl

a a
{

b0

a o
ra')

{) c)

.o
tr

cl
|l)

o.
{.)

,.o

A. O.

a o (l

o
o)

A. a
q-.

lat

o\

rn 6

CI o\

o o

ct\

r-

rtr

(\l ra
c.t

(a

*. (l)

t *. l. o o o o
cl
ql
TA

L.

o
I a

s tr
U

bt cl

GI

ts
(h

GI

c
FI
CA

o
GI

o
Gl

o o
GI

bT

(n

cl

rq EI
tt)

()

L) tI]

a
A

tr{

O{

Or

o o o o A O t) U O ()
GI

tt

14 (a

z o
(n Oi

t1

ah

o o
GI

d
CA

tr a
IA

6l

14 14

o ()

(t)

O{

rd rq

a z ()
14

o o fJ f_)
th

ct)

E o{ tr fr d
M
14

ql

A u
A{

M p{

14

)4 9r M

F ei

tr{ P{

t-

t pi F-

z o

tF tAi tA Fr O{

F F l* Fi A Fr

M t4 M G'

z
F.
Fr

d tr{ A bl o F .o F 5 o{ O{

tD FA

tr{

A A A trr o. P{ A A

(\!

M M

F F o. O{

F A

O{

F-

3 O{

F-r

Oi

F F n F Or A Or t*
A

d li

a
E
cl
GI

o
(r)

o E al 5 -v 6l
ra
CI qt

o o

A
>r

o cl
Gt slt

o "o

>r

GI GI

G'

ct

cl
cl
rt,

u) o

c M
(0

o o
'E' cl

CA

o (l o *.

(l

lt)

tr tr
O{ oa

a
\0

o u d cl (n 6l .o (l c, pa lD E tr o tr o v) (it cl z 14 (! ql o (l N E o *{ .B aD A ah c o o ql rq
CA

cl

(D

cl
6
6d

o g 3 o

&
d "o

cl cl

o
(rt

t a
o a

(! M
E
at

rE

tr

-0) (q

d d
GI

u)
(g

&

& & &

rat

cl
6l

tr ra GI

.o
,.o

a ,t
ql

cl

rn

'i d f{ N
GT

ct

d a

tr

6t co !+ lrl

r-

o\

ol tl \f

\o cr

o\ (>

GI c.) c\I c{

$ r (\l c.l

u0

cl E

o
sl
)i
GI (A

h0 u)

tr

A
a0
GI

|l'

*r lD ll,
a0

a
li

o o a
GI

d J'

It) 6)

.54

&
cl

ct 6l ql

F]
(D

E o o

trr

lD

d
GI

cl

qt

E Fr
(l) {l)

z
;) g
cl
(A

o .tr

a
I
o o ' to
.Y

It

o
qt
I

c
qi

-t 'ct
(l)

tr , o o .$ 3 cl F ql ot)
(D

F.

c cl
L

v,

c)
(D

tr

M
b0

d q

io 6 (D l<i

o cl

a
tn
d,

o
ra.t

'rt q) o E

o cl o () o o
lD

a o[ GI o a0 o o ct (l) g th ttI L tr ll) rtl F v GI E o !t) FA U -v tr o (' tu cl o) b0 o a o o C) qt () cl tr b0 C) (D B a0 GI !E CI .v C) e o tq h o (l) (D M A fri o tr o L) E a F Or F5 a c) b{ b0 o0 c g a sl s c) GI J' o G! c) tt o

(tt

G
ql

c) trl

lt)

ts

o o

c b0 ql L 'o ct 5

F{

(t) cl

o o '5'

ch

U
(.)

a trl

14

&
vt
rc

6l g

b0 6l L FA 14

6l

or

tr{

sl
d)
t4-a

c)

o
r

o o

F]

bo 'a t

GI

C)

o-

o *.

&
o\

o
(l)

O.

a/t

in

g\

rat

\o t\

6t \o t-.

oo

o c.t

o\ t- c{ N

&
o
h d tr
tr!

r3

tr (l)

o o
GI

o o o o
qt
Gt

cl

a
Or

tr
a
6

(l

(l

(t

o
GI

6 (D

() U

o a a o o
{l)

.A

tA CA a r4 F] F] .A

E .E
.A

tr{

U U L)

o o o

q! o o o (, t) .A qt GI C6 cl cl al 6 tr rJ1 rr1 I rq tr g E ts E .A a ah (h o o t) 6 o aD cn (rt = A (J gr O U 14 U tri r{ rrl rd rll r4 U a" A

F]

l.

*{

o o

tr

A
,y o
Pr

GI cl cl o G' 6t GI () al at I o h t, aa b h GI d G' ct ((l GI s .q 6t GI cl o .v (t) () GI o o o cl ct "trt () o "E tr'l tr{ M M M M g v v o M & M M rq \4 GI lt) .(, 6 GI al GI cl ql 13 al GI "o qt GI F g (l) d GI o L li E J d c) Jd 6l u GI E a FA E 'tt cl GI C) tr{ A o ql o (u := (D |D t- iF I* E E ql tr o tn 4t $ A O{ l- F CI A cl Fr A{ t(D GI F 5 F Gt F-l tr{ O. p{ : a F F{ o\ h F F" tC) F F. Or A Pi F- tA p{ \o t{ A Oi A F O. or g{

(l

o{

ct

pq

or

cl
ct vt
-!d

ut

v E
GI

&

o
Cg

g
c)

G,

(A

{)

GI

(l)

o
(.)

aa

& A
tr
A.

cl

\o (\l r- o ot t\ 6l

o tft

; o ll} E o o F o Ft a E 0 ct GI o o o 5 b b{ o N 5 *i rg P{ GI ql c) o E d 'r3 (d G! o s u) 6t F Pr & o a h tu tr o o ch E o .E GI a o t It) IA H E tr !o (D 6l (n lD b0 ll) '(t o GI F o (l) d, m rd hl (5 /, o a :d o A rq z & 4 a (\l ctl .rf in \o f- oo o\ o ol tl $ rr1 \o t- o\ o afi cq tq aa af] ao cn ca crl r<f .+ * :f :f r+ =f, t $ $ r
c GI
cl
CI

.6 o J !
ct

o E
ah

o
ad

63

c)

(A

a & c)

lr p{ "u rd
lD

crl

o (h o o o ll) J a ct L cl o