<%BANNER%>

Project Level Factors Affecting Quality of Construction Projects

Permanent Link: http://ufdc.ufl.edu/UFE0025028/00001

Material Information

Title: Project Level Factors Affecting Quality of Construction Projects
Physical Description: 1 online resource (105 p.)
Language: english
Creator: Bansal, Ankit
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: correlation, cost, management, quality, schedule, scorecard
Building Construction -- Dissertations, Academic -- UF
Genre: Building Construction thesis, M.S.B.C.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Quality management is an important topic in today s construction industry as it has become essential for construction companies to focus on increasing quality performance of construction projects to excel in highly competitive business environment. For the purpose of this research construction quality is defined in terms of project cost growth and project schedule growth. Major objectives of the study: To define a set of factors affecting quality performance of construction projects based on a literature review. To elaborate on each factor affecting quality performance of construction projects in order to have a better understanding of these factors. To establish a correlation between the identified factors and the quality performance of a construction project. To develop a scorecard which can be used to track the status of factors affecting quality performance during a project and thus help project managers be proactive and maintain quality performance. Eight factors namely alliance activeness between owner and contractor, duration of alliance between owner and contractor, rework cost as a % of actual project cost, project team expertise, project team communication, availability of skilled labor, construction productivity, materials availability/cost are analyzed using statistical techniques to determine their relationship with project cost growth and project schedule growth. A scorecard is developed to calculate estimated project cost growth and estimated project schedule growth using a multiple linear regression model. It is also found that low cost growth and schedule growth is associated with high positive impact of these eight factors and there is a possibility of non-linear relationships of these eight factors with project cost growth and project schedule growth.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ankit Bansal.
Thesis: Thesis (M.S.B.C.)--University of Florida, 2009.
Local: Adviser: Ries, Robert J.
Local: Co-adviser: Chini, Abdol R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025028:00001

Permanent Link: http://ufdc.ufl.edu/UFE0025028/00001

Material Information

Title: Project Level Factors Affecting Quality of Construction Projects
Physical Description: 1 online resource (105 p.)
Language: english
Creator: Bansal, Ankit
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: correlation, cost, management, quality, schedule, scorecard
Building Construction -- Dissertations, Academic -- UF
Genre: Building Construction thesis, M.S.B.C.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Quality management is an important topic in today s construction industry as it has become essential for construction companies to focus on increasing quality performance of construction projects to excel in highly competitive business environment. For the purpose of this research construction quality is defined in terms of project cost growth and project schedule growth. Major objectives of the study: To define a set of factors affecting quality performance of construction projects based on a literature review. To elaborate on each factor affecting quality performance of construction projects in order to have a better understanding of these factors. To establish a correlation between the identified factors and the quality performance of a construction project. To develop a scorecard which can be used to track the status of factors affecting quality performance during a project and thus help project managers be proactive and maintain quality performance. Eight factors namely alliance activeness between owner and contractor, duration of alliance between owner and contractor, rework cost as a % of actual project cost, project team expertise, project team communication, availability of skilled labor, construction productivity, materials availability/cost are analyzed using statistical techniques to determine their relationship with project cost growth and project schedule growth. A scorecard is developed to calculate estimated project cost growth and estimated project schedule growth using a multiple linear regression model. It is also found that low cost growth and schedule growth is associated with high positive impact of these eight factors and there is a possibility of non-linear relationships of these eight factors with project cost growth and project schedule growth.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ankit Bansal.
Thesis: Thesis (M.S.B.C.)--University of Florida, 2009.
Local: Adviser: Ries, Robert J.
Local: Co-adviser: Chini, Abdol R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025028:00001


This item has the following downloads:


Full Text

PAGE 1

1 PROJECT LEVEL FACTORS AFFECTING QUALITY OF CONSTRUCTION PROJECTS By ANKIT BANSAL A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN BUILDING CONSTRUCTION UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Ankit Bansal

PAGE 3

3 To my dear parents Renu Bansal and Anil Bansal ; my loving sister Roohi Bansal ; and my friend Kirandeep Kaur who have supported me throughout my graduate study

PAGE 4

4 ACKNOWLEDGMENT I thank God for giving me wisdom and strength to be able to complete this research. I would like to thank Dr. Robert Ries for his kind guidance, support, and encouragement throughout this research. This could not have been completed without his insight, precious comments, and thorough review of my work. I would also like to thank my co -chair Dr. Abdol Chini and my committee member Dr. Douglas Lucas for their valuable suggestions and c ontinuous support. S pecial thanks are given to Kirandeep Kaur who has been great source of motivation and inspiration. I would also like to thank all my colleagues in the b uilding c onstruction p rogram for their valuable assistance. Finally, I would like to express my most p rofound gratitude to my parents, Mrs. Renu Bansal and Mr. Anil Bansal and my sister, Roohi Bansal for giving me their support and encouragement always

PAGE 5

5 TABLE OF CONTENTS Page ACKNOWLEDGMENT ...................................................................................................................... 4 TABLE OF CONTENTS ..................................................................................................................... 5 LIST OF TABLES ................................................................................................................................ 7 LIST OF FIGURES ............................................................................................................................ 12 ABSTRACT ........................................................................................................................................ 14 CHAPTER 1 INTRODUCTION ....................................................................................................................... 16 2 LITERATURE REVIEW ........................................................................................................... 21 2.1 Background ........................................................................................................................ 21 2.2 Discussion of Factors to Be Analyzed ............................................................................. 28 2.2.1 Teamwork .............................................................................................................. 28 2.2.2 Productivity ........................................................................................................... 30 2.2.3 Rework .................................................................................................................. 31 2.2.4 Contractor Selection ............................................................................................. 32 2.2.5 Injuries ................................................................................................................... 34 3 RESEARCH METHODOLOGY AND DATA ANALYSIS ................................................... 35 3.1 Research Structure ............................................................................................................ 35 3.2 Construction Industry Institute Benchmarking & Metrics (CII BM&M) Program ...... 36 3.3 Data Preparation ................................................................................................................ 37 3.4 Statistical Methods Used .................................................................................................. 38 3.5 Project Cost Growth Factor and Project Schedule Growth Factor ................................ 39 3.6 Data Analysis ..................................................................................................................... 39 3.6.1 ANOVA An alysis ................................................................................................. 39 3.6.2 Regression Analysis ............................................................................................. 71 3.6.3 Curve Estimation Analysis ................................................................................... 73 3.6.4 Multiple Linear Regression .................................................................................. 77 3.6.5 Multiple Linear Regression Analysis to Develop Scorecard ............................. 78 4 CONCLUSIONS AND RECOMMENDATIONS ................................................................... 82 4.1 Conclusions ....................................................................................................................... 82 4.1.1 Possible Relationships of Factors with Project Cost Growth ............................ 84 4.1.2 Possible Relationships of Factors with Project Schedule Growth ..................... 84 4.1.3 Scorecard for Estimating Project Cost Growth and Project Schedule Growth 85

PAGE 6

6 4.2 Limitation of Research ...................................................................................................... 85 4.3 Recommendations ............................................................................................................. 86 APPENDIX A SUMMARY OF REGRESSION ANALYSIS FOR CURVE ESTIMATION ....................... 87 B SUMMARY OF INITIAL REGRESSION ANALYSIS WITHOUT DATA GROUPING .. 91 LIST OF REFERENCES ................................................................................................................... 99 BIOGRAPHICAL SKETCH ...........................................................................................................104

PAGE 7

7 LIST OF TABLES Table page 1 1 Factors affecting quality of construction projects ................................................................ 19 1 2 Quality performance criteria and their measurement ........................................................... 20 3 1 CII BM&M survey question for alliance activeness between contractor and owner ........ 40 3 2 Descriptive stat istics for cost growth when analyzing the impact of activeness of alliance between owner and contractor ................................................................................. 41 3 3 Test of homogeneity of variances for cost growth when analyzing the impact of activeness of alliance between owner and contractor .......................................................... 41 3 4 ANOVA table for cost growth when analyzing the impact of activeness of alliance between owner and contractor ............................................................................................... 42 3 5 T test table for cost growth when analyzing the impact of activeness of alliance between owner and contractor ............................................................................................... 42 3 6 Descriptive statistics for schedule growth when analyzing the impact of activeness of alliance between owner and contractor ................................................................................. 42 3 7 Test for homogeneity of variances for schedule growth when analyzing the impact of activeness of alliance between owner and contractor .......................................................... 42 3 8 ANOVA table for schedule growth when analyzing the impact of activeness of alliance between owner and contractor ............................................................................................... 43 3 9 T test table for schedule growth when analyzing the impact of activeness of alliance between owner and contractor ............................................................................................... 43 3 10 CII BM&M survey question for duration of alliance between contractor and owner ....... 45 3 11 Descriptive statistics for cost growth when analyzing the impact of number of years of alliance between owner and contractor ................................................................................. 46 3 12 Test for homogeneity of variances for cost growth when analyzing the impact of number of years of alliance between owner and contractor ................................................ 46 3 13 ANOVA table for cost growth when analyzing the impact of number of years of alliance between owner and contractor ................................................................................. 46 3 14 Descriptive statistics for schedule growth when analyzing the impact of number of years of alliance between owner and contractor .................................................................. 47

PAGE 8

8 3 15 Test for homogeneity of variances for schedule growth when analyzing the impact of number of years of alliance between owner and contractor ................................................ 47 3 16 ANOVA table for schedule growth when analyzing the impact of number of years of alliance between owner and contractor ................................................................................. 47 3 17 T test table for schedule growth when analyzing the impact of number of years of alliance between owner and contractor ................................................................................. 4 7 3 18 Descriptive statistics for cost growth when analyzing the impact of rework ..................... 50 3 19 Test of homogeneity of variances for cost growth when analyzing the impact of rework ..................................................................................................................................... 50 3 20 ANOVA table for cost growth when analyzing the impact of rework ............................... 50 3 21 T test table for cost growth when analyzing the impact of rework ..................................... 51 3 22 Descriptive stati stics for schedule growth when analyzing the impact of rework ............. 51 3 23 Test of homogeneity of variances for schedule growth when analyzing the impact of rework ..................................................................................................................................... 51 3 24 ANOVA table for schedule growth when analyzing the impact of rework ....................... 51 3 25 T test table for schedule growth when analyzing the impact of rework ............................. 52 3 26 CII BM&M survey question for availability of skilled labor .............................................. 53 3 27 Descriptive statistics for cost growth and schedule growth when analyzing the impact of availability of skilled labor ................................................................................................ 55 3 28 Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of availability of skilled labor ............................................................................. 55 3 29 ANOVA table for cost growth and schedule growth when analyzing the impact of availability of skilled labor .................................................................................................... 55 3 30 CII BM&M survey question for materials availability/cost ................................................ 57 3 31 Descriptive statistics for cost growth and schedule growth when analyzing the impact of materials availability/cost .................................................................................................. 59 3 32 Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of materials availability/cost ............................................................................... 59 3 33 ANOVA table for cost growth and schedule growth when analyzing the impact of materials availability/cost ...................................................................................................... 59

PAGE 9

9 3 34 CII BM&M survey question for construction productivity ................................................. 61 3 35 Descriptive statistics for cost growth and schedule growth when analyzing the impact construction productivity ....................................................................................................... 62 3 36 Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of construction productivity ................................................................................ 62 3 37 ANOVA table for cost growth and schedule growth when analyzing the impact of construction productivity ....................................................................................................... 63 3 38 CII BM&M survey question for project team communication ........................................... 65 3 39 Descriptive statistics for cost growth and schedule gro wth when analyzing the impact of project team communication ............................................................................................. 66 3 40 Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of project team communication .......................................................................... 66 3 41 ANOVA table for cost growth and schedule growth when analyzing the impact of project team communication ................................................................................................. 66 3 42 CII BM&M survey que stion for project team expertise ...................................................... 68 3 43 Descriptive statistics for cost growth and schedule growth when analyzing the impact of project team expertise ........................................................................................................ 69 3 44 Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of project team expertise ..................................................................................... 70 3 45 ANOVA table for cost growth and schedule growth when analyzing the impact of project team expertise ............................................................................................................ 70 3 46 Summary of regression analysis ............................................................................................ 72 3 47 Data preparation example for curve estimation analysis ..................................................... 73 3 48 Curve estimation analysis summary ..................................................................................... 74 3 49 Model summary for multiple linear regression .................................................................... 78 3 50 ANOVA table for the multiple linear regression model ...................................................... 78 3 51 Correlation coefficients for the factors in multiple linear regression model ...................... 78 3 52 Model summary for multiple regression analysis for predicting project cost growth ....... 79 3 53 ANOVA table for multiple regression analysis for predicting project cost growth .......... 80

PAGE 10

10 3 54 Correlations for multiple regression analysis for predicting project cost growth .............. 80 3 55 Model summary for multiple regression analysis for predicting project schedule growth ..................................................................................................................................... 80 3 56 ANOVA table for multiple regression analysis for predicting project schedule growth .. 80 3 57 Correlations for multiple regression analysis for predicting project schedule growth ...... 80 A 1 Summary of regression analysis for curve estimation between cost growth and alliance activeness ................................................................................................................................ 87 A 2 Summary of regression analysis for curve estimation between schedule growth and alliance activeness .................................................................................................................. 87 A 3 Summary of regression analysis for curve estimation between cost growth and alliance duration ................................................................................................................................... 87 A 4 Summary of regression analysis for curve estimation between schedule growth and alliance duration ..................................................................................................................... 87 A 5 Summary of regression analysis for curve estimation between cost growth and availability of skilled labor .................................................................................................... 87 A 6 Summary of r egression analysis for curve estimation between schedule growth and availability of skilled labor .................................................................................................... 88 A 7 Summary of regression analysis for curve estimation between cost growth and materials availability .............................................................................................................. 88 A 8 Summary of regression analysis for curve est imation between schedule growth and materials availability .............................................................................................................. 88 A 9 Summary of regression analysis for curve estimation between cost g rowth and project team expertise ......................................................................................................................... 88 A 10 Summary of regression analysis for curve estimation between schedule growth and project tea m expertise ............................................................................................................ 88 A 11 Summary of regression analysis for curve estimation between cost growth and project team communication .............................................................................................................. 89 A 12 Summary of regression analysis for curve estimation between schedule growth and project team communication ................................................................................................. 89 A 13 Summary of regression analysis for curve estimation between cost growth and construction productivity ....................................................................................................... 89

PAGE 11

11 A 14 Summary of regression analysis for curve estimation between schedule growth and construction productivity ....................................................................................................... 89 A 15 Summary of regression analysis for curve estimation between cost growth and rework cost .......................................................................................................................................... 89 A 16 Summary of regression analysis for curve estimation between schedule growth and rework cost .............................................................................................................................. 90 B1 Summary output for regression between cost growth & alliance activeness ..................... 91 B2 Summary output for regression between schedule growth & alliance activeness ............. 91 B3 Summary output for regression between cost growth & alliance duration ........................ 92 B4 Summary output for regression between schedule growth & alliance duration ................ 92 B5 Summary output for regression between cost growth & availability of skilled labor ....... 93 B6 Summary output for regression between schedule growth & availability of skilled labor ......................................................................................................................................... 93 B7 Summary output for regression between cost growth and materials availability/cost ....... 94 B8 Summary output for regression between schedule growth & materials availability/cost ...................................................................................................................... 94 B9 Summary output for regression between cost growth & project team expertise ............... 95 B10 Summary output for regression between schedule growth & project team expertise ........ 95 B11 Summary output for regression between cost growth & project team communication ..... 96 B12 Summary output for regression between schedule growth & project team communication ....................................................................................................................... 96 B13 Summary output for regression between cost growth & construction productivity .......... 97 B14 Summary output for regression between schedule growth & construction productivity .. 97 B15 Summary output for regression between cost growth & rework cost ................................. 98 B16 Summary output for regression between schedule growth & rework cost ......................... 98

PAGE 12

12 LIST OF FIGURES Figure page 3 1 Overview structure for research methodology ..................................................................... 36 3 2 Data preparation example ...................................................................................................... 37 3 3 QQ plot when analyzing the impact of activeness of alliance between owner and contractor ................................................................................................................................ 41 3 4 Box plot when analyzing the impact of activeness of alliance between owner and contractor ................................................................................................................................ 43 3 5 Mean growth comparison when analyzing the impact of activeness of alliance between owner and contractor .............................................................................................................. 44 3 6 QQ plot when analyzing the impact of number of years of alliance between owner and contractor ................................................................................................................................ 46 3 7 Box plot when analyzing the impact of number of years of alliance between owner and contractor ......................................................................................................................... 48 3 8 Mean growth when analyzing the impact of number of years of alliance between owner and contractor .............................................................................................................. 48 3 9 QQ plot when analyzing the impact of rework .................................................................... 50 3 10 Box plot when analyzing the impact of rework ................................................................... 52 3 11 Mean growth comparison when analyzing the impact of rework ....................................... 52 3 12 QQ plot when analyzing the impact of availability of skilled labor ................................... 54 3 13 Box plot when analyzing the impact of availability of skilled labor .................................. 56 3 14 Mean growth comparison when analyzing the impact of availability of skilled labor ...... 56 3 15 QQ plot when analyzing the impact of materials availability/cost ..................................... 58 3 16 Box plot when analyzing the impact of materials availability/cost .................................... 59 3 17 Mean growth comparison when analyzing the impact of materials availability/cost ........ 60 3 18 QQ plot when analyzing the impact of construction productivity ...................................... 62 3 19 Box plot when analyzing the impact of construction productivity ..................................... 63 3 20 Mean growth comparison when analyzing the impact of construction productivity ......... 63

PAGE 13

13 3 21 QQ pl ot when analyzing the impact of project team communication ................................ 65 3 22 Box plot when analyzing the impact of project team communication................................ 67 3 23 Mean growth comparison when analyzing the impact of project team communication ... 67 3 24 QQ plot when analyzing the impact of project team expertise ........................................... 69 3 25 Box plot when analyzing the impact of project team expertise .......................................... 70 3 26 Mean growth comparison when analyzing the impact of project team expertise .............. 71 3 27 Relationship of alliance activeness between contractors and owners with growth factors ...................................................................................................................................... 75 3 28 Relationship of alliance duration between contractors and owners with growth factors .. 75 3 29 Relationship of availability of skilled labor with growth factors ........................................ 75 3 30 Relationship of materials availability/cost with growth factors .......................................... 76 3 31 Relationship of project team expertise with growth factors ................................................ 76 3 32 Relationship of project team communication with growth factors ..................................... 76 3 33 Relationship of construction productivity with growth factors ........................................... 77 3 34 Relationship of rework cost as % of actual project cost with growth factors .................... 77 3 35 Scorecard for estimating project cost growth and project schedule growth ....................... 81

PAGE 14

14 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of S cience in Building Construction PROJECT LEVEL FACTORS AFFECTING QUALITY OF CONSTRUCTION PROJECTS By Ankit Bansal August 2009 Chair: Robert Ries Cochair: Abdol Chini Major: Building Construction Quality management is an important topic in todays construction industry as it has become essential for construction companies to focus on increasing quality performance o f construction projects to excel in highly competitive business environment. For the purpose of this research construction quality is defined in terms of project cost growth and project schedule growth. Major objectives of the study: To define a set of fa ctors affecting quality performance of construction projects based on a literature review. To elaborate on each factor affecting quality performance of construction projects in order to have a better understanding of these factors. To establish a correlation between the identified factors and the quality performance of a construction project. To develop a scorecard which can be used to track the status of factors affecting quality performance during a project and thus help project managers be proac tive and maintain quality performance. Eight factors namely alliance activeness between owner and contractor, duration of alliance between owner and contractor, rework cost as a % of actual project cost, project team expertise, project team communication, availability of skilled labor, construction productivity,

PAGE 15

15 materials availability/cost are analyzed using statistical techniques to determine their relationship with project cost growth and project schedule growth. A scorecard is developed to calculate esti mated project cost growth and estimated project schedule growth using a multiple linear regression model. It is also found that low cost growth and schedule growth is associated with high positive impact of these eight factors and there is a possibility of nonlinear relationships of these eight factors with project cost growth and project schedule growth.

PAGE 16

16 CHAPTER 1 INTRODUCTION Quality management is an important topic in todays construction industry. It has been a fertile area for researchers in several areas including cost of quality, best practices, total quality management and quality performance management systems. Despite of the fact that extensive research has been done in this field, every study has given a different perspective of the subject. There has been a significant improvement over the years regarding the ways of looking at quality management as an important part of construction industry. Today, in such a competitive business environment, it has become immensely important for construction companies to focus on increasing quality performance of construction projects. For an organization it has become essential to identify the factors affecting quality performance of construction projects in order to become the best in the business. It has become increasingly important for project managers to have reliable tools that can help in assessing and controlling the quality performance. It is a complex task to measure the quality performance of a project in terms of success or failure, although it looks simple. Quality is an intangible term and also because of the fact that construction projects involve multidisciplinary teams consisting of architects, designers, project managers, contractors, subcontractors, suppliers, each discipline can have a different definition of success and failure depending upon individual goals and objectives. Therefore, it is important to define quality first in order to assess it on different projects. This holds true for identifying factors that affect the quality performance of projects. Quality performance can be defined in terms of compliance to schedule and budget and customer satisfaction. There exist several factors that are responsible for a project success. These factors if carefully monitored can help in achieving greater success. Similarly there are several factors that

PAGE 17

17 are responsible for failure of projects. These factors if not monitored carefully might prove detrimental for success of the project. Significant improvement is needed to develop a tool that can help control factors which affect project cost and schedule. Major objectives of the study: To define a set of factors affecting quality performance of construction projects based on a literature review. To elaborate on each factor affecting quality performance of construction projects in order to have a better understanding of these factors. To establish a correlation between the identified factors and the quality performance of a construction project. To develop a scorecard which can be used to track the status of factors affecting quality performance during a project and thus help project managers be proactive and maintain quality performance. The role of a Quality Management System is to ensure quality on construction projects. It is necessary to develop a system that can help monitor the key factors influencing construction project quality on a continuous and ongoing basis with respect to time, so that proper decisions can be made at the right time in order to maintain the quality of a construction project Due to the limited research in the field of measurement and analysis in construction projects, there is a relevant need to explore this area in order to identify a set of the most important metrics. Measuring certain variables that affect construction quality helps in forming a benchmark for the range of values these variables should have when an optimal quality level is achieved. It would be very useful to identify those variables which have the best correlation with project quality, in order to control the quality of a project most effectively. Based on the correlation a Quality Scorecard can be developed which can be expressed as a list of the most critical factors that influence construction project quality along with their measurement units so that

PAGE 18

18 these factors can be evaluated during the project execution. Project managers are responsible for the overall success of a construction project, which includes meeting goals related to cost and schedule. Using this tool will help project managers be proactive in an effort to achieve higher quality performance. It will help project managers to take corrective actions if significant deviations occur. Analysis Structure: After collecting the data for Factors affecting quality of construction project and Quality performance criteria, a correlation analysis will be performed between Variables affecting quality and Quality Performance variables. A regression model that best fits the correlation will be obtained and coefficient of determination (R2) will be calculated as a measure of goodness fit to the regression model. In order to accomplish this, the following types of cases will be required in order to compare variable values: Cases where quality was high or good Cases where quality was low or poor After analyzing the project data, and identifying the most important metrics for quality, a tool, i.e. a scorecard, will be developed which will assist project managers by helping them focus on key areas during the project. Construction project quality depends on many factors. A literature review identified the following factors as shown in Table 1 1 that were found to have effect on construction quality. The table shows list of factors that affect the quality of construction projects along with the possible metrics that can be used to measure these factors during construction.

PAGE 19

19 Table 1-1. Factors affecting quality of construction projects Factors Possible Metrics Change Orders % of TIC, Change Order Turnaround Time Communication Workers kept current with design changes, schedule changes, frequency of conflicting instructions. 5 Point Scale: (Always, Often, Sometimes, Rarely, Never) can be used Drug and Alcohol Testing Yes/No, Frequency of testing Employee Commitment/Employee Relations Degree of commitment/collaboration. 3 Point Scale: (High, Medium, Low) can be used Equipment Utilization Cost of Idle hours as a % of TIC, # of idle hours and # of hours equipment was used Job-site Documentation Records maintained in job-site office (Yes/No) Labor Productivity Potential labor hours required on the project and Actual labor hours worked on the project Rework Rework cost as % of TIC Safety # of injuries and # of workers employed on the project Subcontractor Evaluation Cost saving ration, Schedule shortened ration, On time completion frequency, Rework occurrence rate, Quality management execution plan: level of execution, Cooperation in work: level of cooperation. 3 Point Scale: (Excellent, Good, Low) can be used Supplier Quality Management On-time delivery frequency, # of years of experience Teamwork Assessing team work in terms of different elements such as single focus of team, easily accessible information for design and construction, free flow of information, flexibility of team and its responsiveness to changing situations, healthy relationships and respect for each other, accountability 3 Point Scale: (Full achievement, Partial, No) can be used to evaluate each dimension Top Management Involvement Score from CII Structured Interview Worker Training # of hours of training on specific project Use of Corrective Actions Team/Quality Improvement Team on job-site Yes/No, Duration

PAGE 20

20 Quality can be defined in terms of schedule performance, budget performance and level of customer satisfaction as shown in Table 1-2. Table 1-2. Quality performance criteria and their measurement Quality Performance Criteria Criteria Possible Metrics Schedule Performance Ahead On Behind Budget Performance Within Budget Exceeding Budget Level of Customer Satisfaction Low Medium High

PAGE 21

21 CHAPTER 2 LITERATURE REVIEW 2.1 Background Wuellner (1990) presented a checklist that consisted of several parameters to evaluate the performance of a consulting engineering firm. The attributes of a successful project were defined. The four main goals were used in order to design the checklist: It should be cover the key performance criteria comprehensively Should be easy and straightforward to complete Should be useful and easy to complete for project managers without acting as a burden on them Should have the flexibility to be used on different types of projects The critical areas in which checklist measured the performance were: Professional image Quality of design/service Profitability Risk management Conformance to schedule and budget Customer satisfaction Availability of real time information can be useful for project participants in order to manage projects effectively. Russell et al. (1997) conducted a research to establish a process so that project participants could use time dependent variables to predict final project results from start to finish of the project. S-curves were developed for following two project outcomes: Successful: meeting or exceeding budget and schedule expectations Less than successful: not meeting budget and/or schedule expectations of owner Several continuous variables were used as predictor variables and it was found that that prediction power of variables changed with stages of project. Differences between two outcome categories were analyzed using statistical analysis.

PAGE 22

22 Chua et al. (1999) conducted a study to identify critical success factors for different project objectives. Primary importance was given to budget performance. Neural network approach was used to identify eight factors affecting budget performance. The factors identified were: number of organizational levels that were present in between the project manager and the craft workers quantity of detailed design completed at the start of construction number of control meetings that were held during the construction phase of the projects number of updates on budget that were made execution of a constructability plan total turnover of the team amount of money that was spent on controlling the project technical experience of the project manager Realizing the need of a simple and direct measurement for project success that could be used on wide variety of projects based on type and size, Griffith et al. (1999) conducted a research by utilizing data from completed projects and telephone interviews and developed an index that could be used to measure the success of industrial project execution. The index included four variables: Budget attainment Schedule attainment Capacity of design Plant utilization The fact that every construction project is unique, and the increasing complexity of projects makes it highly challenging and difficult to control cost, schedule and quality. Reconstruction projects involve additional factors as compared to a normal building project (Krizek et al. 1996). Cost, Schedule, and quality are highly inter-related and affect one another.

PAGE 23

23 Achieving high excellence or poor performance in one aspect may lead to poor performance of another aspect. Therefore, it is necessary to develop a tool that deals with balanced combination of these aspects to achieve effective performance. McKim et al. (2000) conducted a detailed analysis of the suitability of existing techniques for controlling performance indicators such as cost, schedule, and quality in reconstruction projects. Following criteria were used for calculation of performance variables: Cost Performance Factor CPF (%) = (Total value of change orders / Original contract value) X 100 Schedule Performance Factor SPF (%) = (Total project delay / Original project duration) X 100 Quality Performance a. Estimated rework and/or repair cost b. Number of requests of rework and/or repair c. Number of complaints by the customers that related to noise, dust, smoke, etc. Correlation analysis was done and it was concluded that new construction projects perform much better than reconstruction projects that showed higher schedule over-runs and cost overruns. Two major factors were identified causing schedule and cost over-runs in reconstruction projects: Unexpected site conditions Scope of work changes It was found that by providing cash allowances and by merging the schedule of the facility into the regular construction schedule, cost and schedule over-runs could be reduced. Following problems were identified that were unique for reconstruction projects: Not having proper information about the operating facility Space restrictions for construction Maintaining health and safety of the people occupying the facility Involving more building users

PAGE 24

24 Chan et al. (2001) conducted a study for design and build projects to identify project success factors and examine their relative importance. Factor analysis technique was used to perform the study. Six project success factors identified were: These commitment of the project team, contractors abilities, assessment of risk and liability, customers abilities, needs of endusers, and restrictions put by end-users. Multiple regression analysis yielded that project team commitment, clients competencies, and contractors competencies were important to bring successful project outcome. Iyer et al. (2004) carried out a study to investigate factors affecting cost performance of Indian construction projects. Relative importance index was used to determine relative ranking of attributes. Spearmans rank correlation coefficient was used to compare the ranks of attributes for owners and contractors. Factor analysis yielded that critical success factors were: Competency of the project manager Support by the top management Coordination and leadership skills of the project manager Monitoring and feedback provided by the project members Synchronization among different project members Owners ability and favorable climatic condition Factors adversely affecting the cost performance of projects were identified as: Disagreement between project members Lack of knowledge Lack of cooperation and presence of bad project specific attributes Unfavorable socio economic and climatic condition Not taking decision in time Existence of aggressive competition at tender stage Not enough time to prepare bid The construction industry is widely known for its lack of pace in adoption of new technologies. However, it is seen that this trend has been changing in recent years. Several studies have been conducted in order to identify the effects of using automation and integration technologies. Fergusson (1993) tried to identify a correlation between facility integration and

PAGE 25

25 quality and concluded that there was a strong correlation. Griffis et al. (1995) examined the effects of using three dimensional (3D) computer models on cost, schedule and rework metrics and concluded that use of 3D models helped in cost, schedule and rework reduction. Thomas et al. (2001) examined the effects of design/information technologies by establishing a correlation with project performance in terms of cost growth, schedule growth and safety success and concluded that use of design/information technologies may result in cost savings and schedule reductions. OConnor et al. (2004) conducted a study to examine the associations between technology usage and project performance. The project performance variables that were analyzed consisted of: Project cost performance Project schedule performance Cost success and failure were defined. Similarly, schedule success and failure were defined. Twenty two research hypotheses were analyzed and it was concluded that several technologies may contribute significantly to project performance in terms of cost and schedule success. Technology utilization was found to have a greater association with schedule success as compared to cost success. Job performance has a significant association with project performance. Ireland (2004) suggested that possession of required professional standards by the participants involved in a project can help in reducing costs by up to 10% and project schedule by up to 20%. Cheng et al. (2007) conducted a study to examine the effects of various aspects of job performance on project performance. They attempted to study the association between performers and success of the project. In order to accomplish this, four categories of job performance dimensions were extracted by using exploratory factor analysis. The four job performance categories were treated as independent variables and overall project performance was used as a dependent variable.

PAGE 26

26 Using these variables a hypothesized model was developed. This model was tested using path analysis technique and it was concluded that the task category of job performance had a significant relationship with final project outcomes. Earned value management (EVM) is a technique that helps in predicting final cost of the project and thus helps in project control. With an aim to improve the capability of project managers to make right decisions at right time by providing them with a reliable forecasting method that could help in predicting final cost and duration, Lipke et al. (2008) conducted a study and found the results for both final cost and duration to be satisfactorily reliable for general application of forecasting method. The methods used were EVM, earned schedule (ES), statistical prediction and testing methods. Further analysis indicated that coordination among project participants was the most significant of all the factors having maximum positive influence on cost performance. Bryde (2003) proposed a model for assessment of project management performance. The model proposed six criteria for assessing project management performance. These criteria were based on EFQM business excellence model. It was suggested that this model could be used to improve project management performance by introducing some variation according to an organizations specific needs, goals and objectives. Qureshi et al. (2008) verified this assumption made by Bryde to see whether use of project management performance assessment model bring a change in companys performance or not and further studied which factors of this model had a stronger impact on the project performance. Seven variables were considered for analysis. Six were dependent: PM Leadership, PM Staff, PM Policy & Strategy, PM Partnership & Resources, Project Life Cycle Management Process and Key Performance Indicators and one was independent variable: Project Management Performance. Correlation and Regression analysis

PAGE 27

27 was done on these variables and it was concluded that project management performance assessment model had a potential use as a framework to assess project management performance. The two most common dimensions of project success factors that have been suggested by the researchers and project management community are completion with time and completion within budget. Several studies have been conducted for different developing countries and for different project types which dealt with time and cost performance. Development projects are well known for cost over-runs and inability to comply with schedule (Flyy Bjerg et al. 2003; Matta et al. 2003; Evans 2005). An attempt was made to establish an empirical relationship between time and cost performance and to predict construction time as a function of cost by Bromilow (1974), Kaka et al. (1991), Kumaraswamy et al. (1995), and Chan (2001). Raftery (1994) suggested that poor performance of construction projects is a result of excessive time and cost over-runs. Kaliba et al. (2008) conducted a research and concluded that causes of schedule delay in road construction projects of Zambia were: monetary processes and difficulties faced by contractors and customers, modification of contract, economic problems, materials procurement, drawing changes, problems related to staffing, unavailability of equipment, lack of proper supervision, construction errors, lack of coordination on site, changes in specifications and disputes and strikes by laborers. Ahsan et al. (2009) conducted a study to analyze cost and schedule performance of development projects. The study carried out an in-depth performance examination of various international development projects in terms of cost and time. It was found that projects which took longer than scheduled time for completion experienced cost under-run. The main reasons responsible for schedule delay and cost under-run were identified. Yung et al. (2009) utilized regression models and investigated hard factors affecting construction quality with a conclusion that higher quality is associated with higher power of machinery per

PAGE 28

28 laborer, the use of more plants or machinery per m2 of floor area, properties with larger unit areas, the growth of GDP, the higher labor productivity tender. Quality was defined as the total floor areas of good quality projects as a percentage of total floor area of completed projects in a particular province in a particular year. Standards for Assessment of Quality of Construction and Installation Projects (GBJ300-88) were used to estimate the quality level which was divided into three levels: good, pass and fail. 2.2 Discussion of Factors to Be Analyzed 2.2.1 Teamwork Federle et al. (1993) did a study on applying total quality management to design and construction. Total Quality Management (TQM) in Building Design and Construction workshops were held at Iowa State University. It included breakout sessions, summary sessions, and case study presentations. From the results of the breakout sessions it was found that teamwork played an important role in successful implementation of TQM on construction projects. It was suggested that teamwork should be one of the areas to be covered within the quality awareness training. As one of the strategies for improvement, it was suggested that all firms should aim towards establishing cross-functional process improvement teams that will help in increasing the integration within the company and to benefit from the input of all concerned with a process. It was discussed that team training should be given importance in order to improve quality. Construction industrys distributed approach towards delivering a project and lack of formation of effective, well organized teams has had a negative effect on efficiency of project delivery (Egan 1998; Egan 2002; Evbuomwan et al. 1998). Lack of integration between the involved parties during procurement has also resulted in poor performance (Love et al. 1998). Research done by Latham (1994), Bourn (2000) and Egan (Egan 1998; Egan 2002) has encouraged the industry to shift from its traditional approaches towards the ones that are based

PAGE 29

29 on higher degree of collaboration and integration. Accelerating Change (Egan 2002) provided a very useful and important finding that integration of process and team helps in inducing change that will facilitate the industry to become more successful (Egan 2002). Team integration practices have been mostly used to improve project procurement and project delivery processes (Baiden et al. 2003). However, the amount of success that has been achieved has been lower than expectations, because of existence of adversarial cultures and attitudes (Moore et al. 1999; Moore et al. 2001). Such instances where significant level of integration practices has been achieved in the industry are very less (Vincent et al. 1995; Vyse 2001). Most of the teams that are involved in a construction process, work in order to achieve the goals and objectives which are specific to and defined by their own groups. Achieving overall project success falls behind achievement of individual organizational metrics, thus construction industry has not realized the benefits of teamwork in improving productivity and quality (Glassop 2002; Hayes 2002). Encouraging team integration and bringing together various project parties so that collective strength of all the teams can be used efficiently, can immensely contribute towards the success of the project (Akintoye et al. 2000; Howell 1996; Payne et al. 2003). There is a relevant need for the construction industry to establish a measurement system that provides a dependable assessment of how well team members are working together. By continually measuring team integration using such a tool, quality can be managed in proactive way, rather than having to rectify poor performance after it has occurred. It will help project teams to meet a projects quality requirements at the right cost and on time. Baiden et al. (2006) conducted a study to find out the extent to which successful building and construction managers were able to integrate project delivery teams and also discussed the challenges in achieving team integration. It was concluded that either fully integrated teams are

PAGE 30

30 not necessary for effective team operations within the industry, or that the sector must overcome significant organizational and behavioral barriers if the benefits of integration re to be fully recognized in the future. 2.2.2 Productivity With the expectations and requirements of customers getting higher, the complexity of the construction industry is continuously increasing. Job-site productivity can be influenced by several factors which can be categorized into labor characteristics, project work conditions, non-productive activities (Chris Hendrickson 1998). Some of the project impact factors can be listed as following: Complexity of the project and the total project size Availability of labor on job site Equipment utilization on the job site Contractual regulations that bind the parties involved in the project Local climatic conditions Local cultural characteristics Amount of rework performed Amount of work stoppages Productivity acts as an important contributor towards competitive advantage. There are two perspectives to look at increased productivity in the construction industry: consumers perspective and contractors perspective. From the view point of a consumer, increased productivity helps in lowering the costs, shortening schedules, adds value, and helps in getting better returns on investments whereas from a contractors view point, increased productivity helps in achieving greater customer satisfaction, provides an extra advantage to the organization to in order to be competitive and helps in achieving higher profits (Horner 2001).

PAGE 31

31 More commonly, clients are expecting shorter schedules for more complex projects and due to high competition contractors need to lower down their profit margins while taking minimal risk (McTague 2002). 2.2.3 Rework The construction industry lacks behind in implementing quality management principles and tools and due to this rework occurrence is considered to be unavoidable. Rework affects the construction projects by increasing the probability of increased project schedule, project cost and dissatisfied customers. Measuring rework can be useful for management as they can get more insight of quality management practices followed on the project and can identify areas of improvement within a construction organization. Chauvel et al. (1985) suggested that in order to improve an organizations performance, quality must be measured. Dale et al. (1991) suggested that there is a need to measure quality costs as they can be significant, especially in industries such as construction. Rework has been a persistent problem in some countries with the costs ranging between 12% and 15% of the total project cost (Davis et al. 1989; Neese et al. 1991). According to Taneja (1994), in structural and interior works of projects, the costs of rework can be range from 4% to 12% or average 8% of the total budget. These costs comprise approximately 46% for error in execution, 30% for error in designing and the rest are for the poor quality of material, misunderstanding of drawing, and external factors. The Construction Industry Development Agency (CIDA) in Australia (CIDA 1995) has estimated the direct cost of rework in construction to be greater than 10% of project cost. Thus, if a 10% rework value was applied to the annual turnover of the Australian construction industry in 1996, which was estimated at $43.5 billion per annum (DIST 1998), then the cost of rework can be approximated at $4.3 billion per annum. The lack of attention to quality in construction has meant that quality failures, otherwise known as rework or deviations,

PAGE 32

32 have become inevitable features of the construction process. The industry should start focusing more on finding out ways to minimize rework. Love et al. (1997b) conducted a research on rework costs on two building projects. He found that the rework costs were 2.4% and 3.3% of the total project cost. Importance of measuring the quality cost within which the costs of rework dominate has been stressed upon by Low et al. (1998), Rahman (1993), and Davis et al. (1989). They have particularly recognized a necessity of effective quality cost systems that can help in identifying problem areas that are costing organizations money and thus reducing their profits. In order to improve the quality of projects, it is necessary to identify the causes and cost of construction rework (Love et al. 1999a,b). Due to lack of measurement of rework, the industry, construction managers and other professionals involved do not recognize the amount of rework that actually occurs. Alwi et al. (1999) conducted a study to explore the relationship between the quality of site supervision, expressed as training cost, and the rework cost borne by contractors in high-rise building construction. The study suggested that the quality of site supervision in Indonesia was directly related to the supervisors level of experience gained from formal training and had a strong inverse relation to rework costs. Josephson et al. (2002) conducted a study to identify the causes, costs, and magnitude of rework by analyzing seven construction projects in Sweden. The findings revealed that the costs of rework for the case study projects were 4.4% of the construction values of the observation period, and the time needed to correct them was 7.1% of the total work time. 2.2.4 Contractor Selection Construction industry forms a large segment of the U.S. economy. Every year a wide variety of projects are constructed to meet diverse needs of owners and due to the complexity of construction and the fact that every construction project is unique, there exists the potential for the contractor awarded the contract to fail to fulfill the contract requirements that are essential in

PAGE 33

33 order to complete the project. Contractor failure causes a huge loss to the owner because of following factors: Decrease in the job-site productivity in terms of labor and equipment Increased project costs Schedule delays A well established owner-contractor relationship is also important in order to achieve good project outcomes. The traditional relationship between owner and contractor lacks alignment of objectives. This approach is used on many construction projects. Project participants focus on achieving individual goals with little or no concern for effect on others involved. It eventually leads to poor project outcomes. Therefore it is important to have selection criteria in order to evaluate contractors before signing the contract. Schleifer (1987) identified ten main elements of contractor failure: Increase in the size of the project compared to projects that were done before Unfamiliarity with the geographical area Existence of improper cost accounting systems Poor estimate of contract profitability Unfamiliaritywith type of construction Losing important workforce Promoting employees who were inexperienced to a decision-making position too early Poor cost controls for equipment Unsuccessful billing procedures Errors dealing with transcription due to shift to computerized accounting Due to lack of research that quantitatively analyzed the extent of contractor evaluation performed by owners along with the assessment of impact on final project outcomes, Russell et al. (1992) conducted a study, using the hypothesis: owner-performed contractor evaluation and subsequent monitoring will reduce the chance of contractor failure and concluded that contractor failure causes losses for owners in the form of significant cost and schedule over-runs. It was found that owners in-experience and reluctance to evaluate the contractor, lack of comparing

PAGE 34

34 scheduled performance with actual performance in terms of work done and cost, inappropriate contract documents were the main reasons for contractor failure. 2.2.5 Injuries Safety is one of the major concerns of the US construction industry. As compared to other industries, construction industry is considered to have the maximum accident rate. In the year 2004, 1224 fatal occupational injuries and 421,400 nonfatal injuries and illnesses were reported in construction and the incidence rates for nonfatal injuries and illnesses in construction were higher than those in all private industries by 1.6 per 100 full-time equivalent workers. In todays highly competitive scenario, the cost of construction injuries can prove to be immensely harmful for companies. For instance, a company that operates at a 4% profit margin would have to increase contract prices by $400,000 to pay for a $16,000 injury, such as the amputation of a finger. Ahmed et al. (2006) estimated the total cost of fatalities and injuries/illnesses including days away from work in the construction industry and concluded that the total cost of occupational fatal and non-fatal accidents in 2004 was $49.7 billion representing 10.72% of the total turnover of the construction industry. It was indicated that the fatality and injury costs consume a major portion of the construction economy. Therefore it is important for construction companies to focus on implementing strategies that are effective in reducing injuries and increase worker safety on projects. This approach will help in achieving significant cost savings and will also improve worker productivity on jobsites.

PAGE 35

35 CHAPTER 3 RESEARCH METHODOLOGY AND DATA ANALYSIS This chapter elaborates on the research process in detail. Firstly, the overview structure of the research is described using a diagram. Secondly, the data collection process of the research is explained and a brief introduction about Construction Industry Institute (CII) Benchmarking and Metrics (BM&M) database and questionnaire is given. Then, the process followed for the preparation of data for analysis and the statistical methods used for conducting analysis are discussed. 3.1 Research Structure Figure 3-1 briefly summarizes the process followed to fulfill the research objectives. The first step was to conduct extensive literature review to identify the factors affecting quality of construction projects. An extensive literature review yielded a list of factors listed in chapter 1 that affect quality of construction projects. A list of possible metrics that could be used by organizations to measure these factors on construction projects was defined for each factor. A combined list of factors along with their possible metrics was sent to CII in order to get the data. CII BM&M questionnaire was received in response to the data request. Questions that addressed the listed factors were extracted from the questionnaire. CII was then requested to provide data for these selected questions. The data received from CII was in excel sheet format and was then refined removing the unusable data. This data was further used to perform statistical data analysis. Finally, based on the analysis conclusions & recommendations were made. CII database consists of data for building projects, heavy industrial projects, light industrial projects and infrastructure projects. The CII BM&M program helps the members to compare their capital and maintenance projects with other companies that have the best performance. CII mainly deals with capital facilities industry.

PAGE 36

36 Figure 3-1. Overview structure for research methodology 3.2 Construction Industry Institute Benchmarking & Metrics (CII BM&M) Program The data used in the research was collected by the CII BM&M program. The database is populated by CII member companies responding to the CII BM&M questionnaire by submitting data from actual projects. Companies select employees to act as Benchmarking Associates. These Benchmarking Associates are offered training sessions by CII in order to familiarize them with metric definitions and the BM&M questionnaire. This is important in as it helps in gathering consistent and accurate data. These Benchmarking Associates provide training to project managers that actually complete the CII BM&M questionnaire. The questionnaire has underwent a series of revisions and has been constantly updated from Version 1 with data collected in the year 1996, through the newest version 7 consisting of 2002/2003 data. CII BM&M started with paper version of the questionnaire and gradually upgraded to electronic and web questionnaires in 1999. Benchmarking Associates and CII staff work together in order to btn fbrbbb bbbn fb bfb rbb t

PAGE 37

37 minimize ambiguous data and to increase consistency of data. After this validation process is accomplished, the data is added to the BM&M database. 3.3 Data Preparation Microsoft Excel was used to prepare data for the analysis. Figure 3-2. Data preparation example For example, as shown in Figure 3-2 above, column A represents project id, column B represents project cost which is same as actual project cost. Column C represents activeness of alliance between owners & contractors. This is one of the eight factors used in the research. Column D represents actual cost of the project, column E represents budgeted cost of the project, column F, G, H, I represent actual start, actual finish, planned start and planned finish dates for the project respectively. There were in all eight factors that were tested for

PAGE 38

38 relationship with project cost growth and project schedule growth. The data preparation was same for all the factors as shown in the above example and was done in separate worksheets. Each worksheet was sorted in such a way that there was no blank cell. Rows having negative numbers or undefined numbers were deleted. 3.4 Statistical Methods Used Microsoft excel was used to perform correlation & regression analysis for the factors to see if there was a significant degree of correlation between factors affecting construction quality and project cost growth and schedule growth. SPSS Statistics 17.0 was used to perform one-way analysis of variance (ANOVA). T-test assuming unequal variances was used when ANOVAs equal variance assumption was violated. The one-way ANOVA technique is basically used to test differences between means of groups. The null hypothesis (Ho) for this test is that all means of several population groups are equal. On the other hand, the alternative hypothesis (Ha) is that at least one mean of the groups is different from others. The null hypothesis of equal means is tested to see if the resulting p-value is sufficiently small. If not, the analysis stops due to the lack of evidence to reject the null hypothesis. However, if the p-value is small enough to reject the null hypothesis, the alternative hypothesis is accepted. That is, all means are not equal. A p-value is the probability of wrongly rejecting the null hypothesis if it is in fact true. There are two underlying assumptions for this test: (1) the normality of the populations, and (2) the equal variance of the populations. T-test assuming unequal variances was used when equal variance assumption was violated. The level of significance for the one-way ANOVA test is 0.05. The correlation coefficient equals 1 for a perfect positive correlation and -1 for a perfect negative correlation. When the correlation is not perfect, the coefficient lies between -1 and 1. Correlation and regression analysis was performed with null hypothesis (Ho) being that there

PAGE 39

39 was no correlation between the factors affecting construction quality and project outcomes and alternative hypothesis (Ha) being that there was a significant correlation between the factors affecting construction quality and project outcomes. The level of significance for this test is 0.05. 3.5 Project Cost Growth Factor and Project Schedule Growth Factor For this research quality was defined in terms of cost and schedule and the following formulae were used in order to calculate project cost growth and project schedule growth. btnbfr bbbnbbbn bbbnb brbfr b bbbbbb bbb 3.6 Data Analysis Following eight factors affecting quality of construction projects were analyzed based on the selected factors after the conjunction between literature review and CII BM&M questionnaire: 1. Activeness of alliance between owner and contractor 2. Duration of alliance between owner and contractor 3. Availability of skilled labor 4. Materials availability/cost 5. Construction productivity 6. Rework cost 7. Project team communication 8. Project team expertise 3.6.1 ANOVA Analysis Activeness of alliance between owner and contractor: Alliance activeness between contractor and owner involves degree of cooperation and coordination between the two parties in context to working relationships.The survey question in the questionnaire was stated as, to what

PAGE 40

40 extent was this alliance with the primary contractor an active alliance versus just an alliance on paper? Table 3-1. CII BM&M survey question for alliance activeness between contractor and owner Essentially an Alliance on paper A Moderately Active Alliance A Very Active Alliance Dont Know 1 2 3 4 5 6 7 8 The following hypothesis was used for the analysis in terms of cost growth: Ho: There is no difference between mean project cost growth when degree of alliance is high and when it is not. Ha: There is a significant difference between mean project cost growth when degree of alliance is high and when it is not. The following hypothesis was used for the analysis in terms of schedule growth: Ho: There is no significant difference between mean project schedule growth when degree of alliance is high and when it is not. Ha: There is a significant difference between mean project schedule growth when degree of alliance is high and when it is not. For analysis, scale of 5 to 7 was considered to be as highly active alliance and scale of 1 to 4 was considered to be less than highly active alliance. Data was combined into two groups. Highly active alliance was considered group 1 and less than highly active alliance was considered group 2. In order to check the normality of data Quantile-Quantile (QQ) plots were generated for project cost growth and project schedule growth. These plots show quantiles of the scores on the horizontal axis and the expected normal scores on the vertical axis. This graph yields a straight line and deviation of points from the straight line shows departure from normality

PAGE 41

41 QQ plots for cost growth and schedule growth drawn to check normality of data can be seen in Figure 3-3. A B Figure 3-3. QQ plot when analyzing the impact of activeness of alliance between owner and contractor. A) Cost growth. B) Schedule growth. One way ANOVA analysis was performed in order to compare the means of project cost growth for highly active alliance and less than highly active alliance. Descriptive statistics and ANOVA table for the cost growth can be seen in Table 3-2 and Table 3-4. Table 3-2. Descriptive statistics for cost growth when analyzing the impact of activeness of alliance between owner and contractor N Mean 1.00000 28 -.0499291 2.00000 13 .0805286 Total 41 -.0085645 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-3, there was a significant difference between variances for cost growth for the two groups i.e. highly active alliance and less than highly active alliance. Table 3-3. Test of homogeneity of variances for cost growth when analyzing the impact of activeness of alliance between owner and contractor Levene Statistic df1 df2 Sig. 6.158 1 39 .018

PAGE 42

42 Table 3-4. ANOVA table for cost growth when analyzing the impact of activeness of alliance between owner and contractor Sum of Squares df Mean Square F Sig. Between Groups .151 1 .151 12.729 .001 Within Groups .463 39 .012 Total .614 40 As it was found from Levenes statistic test that variances for cost growth for the two groups were significantly different, T-test (equal variances not assumed) was used for comparison of means. Table 3-5. T-test table for cost growth when analyzing the impact of activeness of alliance between owner and contractor Sig. (2-tailed) Equal variances assumed .001 Equal variances not assumed .010 One way ANOVA analysis was performed in order to compare the means of project schedule growth for highly active alliance and less than highly active alliance. Descriptive statistics and ANOVA table for the cost growth can be seen in Table 3-6 and Table 3-8. Table 3-6. Descriptive statistics for schedule growth when analyzing the impact of activeness of alliance between owner and contractor N Mean 1.00000 27 .0578904 2.00000 12 .1714866 Total 39 .0928431 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-7, there was a significant difference between variances for schedule growth for the two groups i.e. highly active alliance and less than highly active alliance. Table 3-7. Test for homogeneity of variances for schedule growth when analyzing the impact of activeness of alliance between owner and contractor Levene Statistic df1 df2 Sig. 9.800 1 37 .003

PAGE 43

43 Table 3-8. ANOVA table for schedule growth when analyzing the impact of activeness of alliance between owner and contractor Sum of Squares df Mean Square F Sig. Between Groups .107 1 .107 4.033 .052 Within Groups .984 37 .027 Total 1.091 38 As it was found from Levenes statistic test that variances for schedule growth for the two groups were significantly different, T-test (equal variances not assumed) was used for comparison of means. Table 3-9. T-test table for schedule growth when analyzing the impact of activeness of alliance between owner and contractor Sig. (2-tailed) Equal variances assumed .052 Equal variances not assumed .136 Box plots for cost growth and schedule growth when analyzing the impact of activeness of alliance between owner and contractor can be seen in Figure 3-4. A B Figure 3-4. Box plot when analyzing the impact of activeness of alliance between owner and contractor. A) Cost growth. B) Schedule growth. Mean cost growth and schedule growth comparison when analyzing the impact of activeness of alliance between owner and contractor can be seen in Figure 3-5.

PAGE 44

44 A B Figure 3-5. Mean growth comparison when analyzing the impact of activeness of alliance between owner and contractor. A) Cost growth. B) Sschedule growth. Conclusion: Mean project cost growth for highly active alliance was different than mean project cost growth for less than highly active alliance at 95% significance level (p=0.010). Mean project cost growth factor for highly active alliance was found to be -5% and mean project cost growth factor for less than highly active alliance was found to be +8%. Mean project schedule growth for highly active alliance was not statistically significantly different than mean project schedule growth for less than highly active alliance (p=0.136). However, mean project schedule growth factor for highly active alliance was found to be +6% and mean project schedule growth factor for less than highly active alliance was found to be +17%. The reason for non-significant result for schedule growth could be that the measure of alliance activeness between the owner and the contracter is a one time criteria and is usually undertaken before the project starts, so this factor might not be affecting the project schedule growth on a continuous basis.

PAGE 45

45 Duration of alliance between owner and contractor: Duration of alliance between contractor and owner deals with the length of time of relationship between the two parties. The survey question in the questionnaire was stated as, how long have you had this alliance with the primary contractor? Table 3-10. CII BM&M survey question for duration of alliance between contractor and owner 1 One year or less 2 One to three years 3 Three to five years 4 More than five years 5 Dont Know The following hypothesis was used for the analysis in terms of cost growth: Ho: There is no difference between mean project cost growth when number of years of alliance is high and when it is not. Ha: There is a significant difference between mean project cost growth when number of years of alliance is high and when it is not. The following hypothesis was used for the analysis in terms of schedule growth: Ho: There is no significant difference between mean project schedule growth when number of years of alliance is high and when it is not. Ha: There is a significant difference between mean project schedule growth when number of years of alliance is high and when it is not. For analysis, alliance of more than 5 years was considered to be as highly active alliance and less than 5 years was considered to be less than highly active alliance. Data was combined into two groups. Alliance of more than 5 years was considered to be group1 and less than or equal to 5 years was considered to be group 2. QQ plots for cost growth and schedule growth drawn to check normality of data can be seen in Figure 3-6.

PAGE 46

46 A B Figure 3-6. QQ plot when analyzing the impact of number of years of alliance between owner and contractor. A) Cost growth. B) Schedule growth. One way ANOVA analysis was performed in order to compare the means of project cost growth for alliance of more than 5 years and alliance of less than or equal to 5 years. Descriptive statistics and ANOVA table for cost growth can be seen in Table 3-11 and Table 3-13. Table 3-11. Descriptive statistics for cost growth when analyzing the impact of number of years of alliance between owner and contractor N Mean 1.00000 21 -.0101748 2.00000 16 -.0080803 Total 37 -.0092691 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-12, the variances were homogeneous (p=0.112). Table 3-12. Test for homogeneity of variances for cost growth when analyzing the impact of number of years of alliance between owner and contractor Levene Statistic df1 df2 Sig. 2.660 1 35 .112 Table 3-13. ANOVA table for cost growth when analyzing the impact of number of years of alliance between owner and contractor Sum of Squares df Mean Square F Sig. Between Groups .000 1 .000 .003 .960 Within Groups .556 35 .016 Total .556 36

PAGE 47

47 One way ANOVA analysis was performed in order to compare the means of project schedule growth for alliance of more than 5 years and alliance of less than or equal to 5 years. Descriptive statistics and ANOVA table for cost growth can be seen in Table 3-14 and Table 316. Table 3-14. Descriptive statistics for schedule growth when analyzing the impact of number of years of alliance between owner and contractor N Mean 1.00000 20 .0505359 2.00000 16 .1165642 Total 36 .0798818 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-15, there was a significant difference between variances for schedule growth for the two groups i.e. alliance of more than 5 years and alliance of less than or equal to 5 years (p=0.035). Table 3-15. Test for homogeneity of variances for schedule growth when analyzing the impact of number of years of alliance between owner and contractor Levene Statistic df1 df2 Sig. 4.816 1 34 .035 Table 3-16. ANOVA table for schedule growth when analyzing the impact of number of years of alliance between owner and contractor Sum of Squares df Mean Square F Sig. Between Groups .039 1 .039 1.667 .205 Within Groups .791 34 .023 Total .829 35 As it was found from Levenes statistic test that variances for schedule growth for the two groups were significantly different, T-test (equal variances not assumed) was used for comparison of means. Table 3-17. T-test table for schedule growth when analyzing the impact of number of years of alliance between owner and contractor Sig. (2-tailed) Equal variances assumed .205 Equal variances not assumed .237

PAGE 48

48 Box plots for cost growth and schedule growth when analyzing the impact of number of years of alliance between owner and contractor can be seen in Figure 3-7. A B Figure 3-7. Box plot when analyzing the impact of number of years of alliance between owner and contractor. A) Cost growth. B) Schedule growth. Mean cost growth and schedule growth comparison when analyzing the impact of number of years of alliance between owner and contractor can be seen in Figure 3-8. A B Figure 3-8. Mean growth when analyzing the impact of number of years of alliance between owner and contractor. A) Cost growth. B) Schedule growth. Conclusion: Mean project cost growth for more than 5 years of alliance between owner and contractor was almost same as mean project cost growth for less than or equal to 5 years of alliance between owner and contractor. Mean project cost growth factor for more than 5 years of alliance between owner and contractor was found to be -1% and mean project cost growth factor for less than or equal to 5 years of alliance between owner and contractor was found to be -0.8%. The difference was not statistically significant (p=0.960).

PAGE 49

49 Mean project schedule growth for more than 5 years of alliance between owner and contractor was not statistically significantly different from mean project schedule growth for less than or equal to 5 years of alliance between owner and contractor (p=0.237). However, mean project schedule growth factor for more than 5 years of alliance between owner and contractor was found to be +5% and mean project cost growth factor for less than or equal to 5 years of alliance between owner and contractor was found to be +11.6%. Rework cost as a % of actual project cost: Direct rework cost is defined by CII as the total direct cost of field rework regardless of initiating cause. The survey question in the questionnaire was stated as, what is the direct cost of field rework? The following hypothesis was used for the analysis in terms of cost growth: Ho: There is no difference between mean project cost growth when the rework cost was high and when the rework cost was low. Ha: There is a significant difference between mean project cost growth when the rework cost was high and when the rework cost was low. The following hypothesis was used for the analysis in terms of schedule growth: Ho: There is no difference between mean project schedule growth when the rework cost was high and when the rework cost was low. Ha: There is a significant difference between mean project schedule growth when the rework cost was high and when the rework cost was low. For analysis, rework cost of more than 2% of actual project cost was considered to be as high rework cost and rework cost of less than 2% of actual project cost was considered to be as low rework cost. Data was combined in two groups. Rework cost of less than 2% of actual project cost was considered group 1 and rework cost of more than 2% of actual project was considered group 2. QQ plots for cost growth and schedule growth drawn to check normality of data can be seen in Figure 3-9.

PAGE 50

50 A B Figure 3-9. QQ plot when analyzing the impact of rework. A) Cost growth. B) Schedule growth. One way ANOVA analysis was performed in order to compare the means of project cost growth for high rework cost and low rework cost. Descriptive statistics and ANOVA table can be seen in Table 3-18 and Table 3-20. Table 3-18. Descriptive statistics for cost growth when analyzing the impact of rework N Mean 1.00000 149 -.0075894 2.00000 70 .0626712 Total 219 .0148683 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-19, there was a significant difference between variances for cost growth for the two groups i.e. high rework cost and low rework cost. Table 3-19. Test of homogeneity of variances for cost growth when analyzing the impact of rework Levene Statistic df1 df2 Sig. 5.445 1 217 .021 Table 3-20. ANOVA table for cost growth when analyzing the impact of rework Sum of Squares df Mean Square F Sig. Between Groups .235 1 .235 6.527 .011 Within Groups 7.817 217 .036 Total 8.052 218

PAGE 51

51 As it was found from Levenes statistic test that variances for cost growth for the two groups were significantly different, T-test (equal variances not assumed) was used for comparison of means. Table 3-21. T-test table for cost growth when analyzing the impact of rework Sig. (2-tailed) Equal variances assumed .011 Equal variances not assumed .036 One way ANOVA analysis was performed in order to compare the means of project schedule growth for high rework cost and low rework cost. Descriptive statistics and ANOVA table can be seen in Table 3-22 and Table 3-24. Table 3-22. Descriptive statistics for schedule growth when analyzing the impact of rework N Mean 1.00000 70 .0885722 2.00000 29 .2642960 Total 99 .1400468 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-23, there was a significant difference between variances for schedule growth for the two groups i.e. high rework cost and low rework cost. Table 3-23. Test of homogeneity of variances for schedule growth when analyzing the impact of rework Levene Statistic df1 df2 Sig. 10.878 1 97 .001 Table 3-24. ANOVA table for schedule growth when analyzing the impact of rework Sum of Squares df Mean Square F Sig. Between Groups .633 1 .633 3.712 .057 Within Groups 16.546 97 .171 Total 17.179 98 As it was found from Levenes statistic test that variances for schedule growth for the two groups were significantly different, T-test (equal variances not assumed) was used for comparison of means.

PAGE 52

52 Table 3-25. T-test table for schedule growth when analyzing the impact of rework Sig. (2-tailed) Equal variances assumed .057 Equal variances not assumed .196 Box plots for cost growth and schedule growth when analyzing the impact of rework can be seen in Figure 3-10. A B Figure 3-10. Box plot when analyzing the impact of rework. A) Cost growth. B) Schedule growth. Mean cost growth and schedule growth comparison when analyzing the impact rework can be seen in Figure 3-11. A B Figure 3-11. Mean growth comparison when analyzing the impact of rework. A) Cost growth. B) Schedule growth.

PAGE 53

53 Conclusion: Mean project cost growth for rework cost less than 2% of actual project cost was different than mean project cost growth for rework cost more than 2% of actual project cost at 95% significance level (p=0.036). Mean project cost growth factor for rework cost less than 2% of actual project cost was found to be -0.7% and mean project cost growth factor for rework cost more than 2% of actual project cost was found to be +6.2%. Mean project schedule growth for rework cost less than 2% of actual project cost was not statistically significantly different than mean project schedule growth for rework cost more than 2% of actual project cost (p=0.196). However, mean project schedule growth factor for rework cost less than 2% of actual project cost was found to be +8.8% and mean project schedule growth factor for rework cost more than 2% of actual project cost was found to be +26.4%. Availability of skilled labor: Availability of skilled labor in the local market that can be hired to perform tasks. The survey question in the questionnaire was stated as, using a scale from -5 to +5, where -5 means an extremely negative impact compared to what was expected or planned and +5 means an extremely positive impact compared to what was expected or planned, please indicate the extent to which availability of skilled labor had a net positive impact, a net negative impact, or was essentially as planned? Table 3-26. CII BM&M survey question for availability of skilled labor Extremely Negative Impact As Planned Extremely Positive Impact -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 The following hypothesis was used for the analysis in terms of cost growth: Ho: There is no difference between mean project cost growth when the availability of skilled labor has highly positive impact and when availability of skilled labor has less than highly positive impact.

PAGE 54

54 Ha: There is a significant difference between mean project cost growth when the availability of skilled labor has highly positive impact and when availability of skilled labor has less than highly positive impact. The following hypothesis was used for the analysis in terms of schedule growth: Ho: There is no difference between mean project schedule growth when the availability of skilled labor has highly positive impact and when availability of skilled labor has less than highly positive impact. Ha: There is a significant difference between mean project schedule growth when the availability of skilled labor has highly positive impact and when availability of skilled labor has less than highly positive impact. For analysis, scale of 0 to +5 was considered to be as highly positive impact and scale of -5 to -1 was considered to be less than highly positive impact. Data was combined in two groups. Highly positive impact was considered group 1 and less than highly positive impact was considered group 2. QQ plots for cost growth and schedule growth drawn to check normality of data can be seen in Figure 3-12. A B Figure 3-12. QQ plot when analyzing the impact of availability of skilled labor. A) Cost growth. B) Schedule growth. One way ANOVA analysis was performed in order to compare the means of project cost growth and schedule growth for highly positive impact and less than highly positive impact. Descriptive statistics and ANOVA table can be seen in Table 3-27 and Table 3-29.

PAGE 55

55 Table 3-27. Descriptive statistics for cost growth and schedule growth when analyzing the impact of availability of skilled labor N Mean Cost_growth 1.00000 34 -.0639750 2.00000 13 -.0158779 Total 47 -.0506716 Schedule_growth 1.00000 32 .0082469 2.00000 13 .0283528 Total 45 .0140552 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-28, the variances were homogeneous. Table 3-28. Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of availability of skilled labor Levene Statistic df1 df2 Sig. Cost_growth .056 1 45 .814 Schedule_growth .000 1 43 .986 Table 3-29. ANOVA table for cost growth and schedule growth when analyzing the impact of availability of skilled labor Sum of Squares df Mean Square F Sig. Cost_growth Between Groups .022 1 .022 3.070 .087 Within Groups .319 45 .007 Total .341 46 Schedule_grow th Between Groups .004 1 .004 .362 .551 Within Groups .444 43 .010 Total .448 44 Box plots and mean cost growth and schedule growth graphs were generated in order to see the significant differences between high vs low impact of availability of skilled labor. Box plots for cost growth and schedule growth when analyzing the impact of availability of skilled labor can be seen in Figure 3-13. The box plot for the cost growth indicates that there is a statistical significant difference between high impact and low impact of availability of skilled labor but the schedule growth box plot does not indicate a significant difference.

PAGE 56

56 A B Figure 3-13. Box plot when analyzing the impact of availability of skilled labor. A) Cost growth. B) Schedule growth. It was seen that there was a difference of magnitude in project cost growth and project schedule growth when comparing high vs low impact of availability of skilled labor. Mean cost growth and schedule growth comparison when analyzing the impact of availability of skilled labor can be seen in Figure 3-14. A B Figure 3-14. Mean growth comparison when analyzing the impact of availability of skilled labor. A) Cost growth. B) Schedule growth. Conclusion: Mean project cost growth for highly positive impact of availability of skilled labor was statistically significantly different than mean project cost growth for less than highly positive impact of availability of skilled labor at 90% significance level (p=0.087). Mean project cost growth factor for highly positive impact of availability of skilled labor was found to be -6%

PAGE 57

57 and mean project cost growth factor for less than highly positive impact of availability of skilled labor was found to be -1.6%. Mean project schedule growth for highly positive impact of availability of skilled labor was not statistically significantly different than mean project schedule growth for less than highly positive impact of availability of skilled labor (p=0.551). However, mean project schedule growth factor for highly positive impact of availability of skilled labor was found to be +0.8% and mean project schedule growth factor for less than highly positive impact of availability of skilled labor was found to be +2.8%. Materials availability/cost: Material availability/cost involves on time delivery of construction materials on the project site, on time availability of the construction materials for use by workers on site and availability of materials in the local market. The survey question in the questionnaire was stated as, using a scale from -5 to +5, where -5 means an extremely negative impact compared to what was expected or planned and +5 means an extremely positive impact compared to what was expected or planned, please indicate the extent to which materials availability/cost had a net positive impact, a net negative impact, or was essentially as planned? Table 3-30. CII BM&M survey question for materials availability/cost Extremely Negative Impact As Planned Extremely Positive Impact -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 The following hypothesis was used for the analysis in terms of cost growth Ho: There is no difference between mean project cost growth when the availability of material has highly positive impact and when availability of material has less than highly positive impact. Ha: There is a significant difference between mean project cost growth when the availability of material has highly positive impact and when availability of material has less than highly positive impact.

PAGE 58

58 The following hypothesis was used for the analysis in terms of schedule growth Ho: There is no difference between mean project schedule growth when the availability of material has highly positive impact and when availability of material has less than highly positive impact. Ha: There is a significant difference between mean project schedule growth when the availability of material has highly positive impact and when availability of material has less than highly positive impact. For analysis, a scale of 0 to +5 was considered to be as highly positive impact and scale of -5 to -1 was considered to be less than highly positive impact. Data was combined in two groups. Highly positive impact was considered group 1 and less than highly positive impact was considered group 2. Looking at the QQ plots it was concluded that data was fairly normal both for project cost growth and project schedule growth. QQ plots for cost growth and schedule growth drawn to check normality of data can be seen in Figure 3-15. A B Figure 3-15. QQ plot when analyzing the impact of materials availability/cost. A) Cost growth. B) Schedule growth. One way ANOVA analysis was performed in order to compare the means of project cost growth and schedule growth for highly positive impact and less than highly positive impact. Descriptive statistics and ANOVA table can be seen in Table 3-31 and Table 3-33.

PAGE 59

59 Table 3-31. Descriptive statistics for cost growth and schedule growth when analyzing the impact of materials availability/cost N Mean Cost_growth 1.00000 31 -.0585777 2.00000 18 .0071617 Total 49 -.0344286 Schedule_growth 1.00000 29 .0078616 2.00000 18 .0323134 Total 47 .0172261 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-32, the variances were homogeneous. Table 3-32. Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of materials availability/cost Levene Statistic df1 df2 Sig. Cost_growth .844 1 47 .363 Schedule_growth .567 1 45 .455 Table 3-33. ANOVA table for cost growth and schedule growth when analyzing the impact of materials availability/cost Sum of Squares df Mean Square F Sig. Cost_growth Between Groups .049 1 .049 3.872 .055 Within Groups .597 47 .013 Total .647 48 Schedule_grow th Between Groups .007 1 .007 .661 .420 Within Groups .452 45 .010 Total .459 46 Box plots when analyzing the impact of materials availability can be seen in Figure 3-16. A B Figure 3-16. Box plot when analyzing the impact of materials availability/cost. A) Cost growth. B) Schedule growth.

PAGE 60

60 Mean cost growth and schedule growth comparison when analyzing the impact of materials availability/cost can be seen in Figure 3-17. A B Figure 3-17. Mean growth comparison when analyzing the impact of materials availability/cost. A) Cost growth. B) Schedule growth. Conclusion: Mean project cost growth for highly positive impact of materials availability/cost was statistically significantly different than mean project cost growth for less than highly positive impact of materials availability/cost at 90% significance level (p=0.055). Mean project cost growth factor for highly positive impact of materials availability/cost was found to be -5.9% and mean project cost growth factor for less than highly positive impact of materials availability/cost was found to be +0.7%. Mean project schedule growth for highly positive impact of materials availability/cost was not statistically significantly different than mean project schedule growth for less than highly positive impact of materials availability/cost (p=0.420). However, mean project schedule growth factor for highly positive impact of materials availability/cost was found to be +0.8% and mean project schedule growth factor for less than highly positive impact of materials availability/cost was found to be +3%. Construction productivity: Construction productivity is defined as number of actual work hours required to perform the appropriate units of work. The survey question in the questionnaire was stated as, using a scale from -5 to +5, where -5 means an extremely negative

PAGE 61

61 impact compared to what was expected or planned and +5 means an extremely positive impact compared to what was expected or planned, please indicate the extent to which construction productivity had a net positive impact, a net negative impact, or was essentially as planned? Table 3-34. CII BM&M survey question for construction productivity Extremely Negative Impact As Planned Extremely Positive Impact -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 The following hypothesis was used for the analysis in terms of cost growth: Ho: There is no difference between mean project cost growth when the construction productivity has highly positive impact and when construction productivity has less than highly positive impact. Ha: There is a significant difference between mean project cost growth when the construction productivity has highly positive impact and when construction productivity has less than highly positive impact. The following hypothesis was used for the analysis in terms of schedule growth: Ho: There is no difference between mean project schedule growth when the construction productivity has highly positive impact and when construction productivity has less than highly positive impact. Ha: There is a significant difference between mean project schedule growth when the construction productivity has highly positive impact and when construction productivity has less than highly positive impact. For analysis, scale of 0 to +5 was considered to be as highly positive impact and scale of -5 to -1 was considered to be less than highly positive impact. Data was combined in two groups. Highly positive impact was considered group 1 and less than highly positive impact was considered group 2. QQ plots for cost growth and schedule growth drawn to check normality of data can be seen in Figure 3-18.

PAGE 62

62 A B Figure 3-18. QQ plot when analyzing the impact of construction productivity. A) Cost growth. B) Schedule growth. One way ANOVA analysis was performed in order to compare the means of project cost growth and schedule growth for highly positive impact and less than highly positive impact. Descriptive statistics and ANOVA table can be seen in Table 3-34 and Table 3-36. Table 3-35. Descriptive statistics for cost growth and schedule growth when analyzing the impact construction productivity N Mean Cost_growth 1.00000 32 -.0591003 2.00000 17 -.0105478 Total 49 -.0422556 Schedule_growth 1.00000 31 .0091350 2.00000 17 .0344391 Total 48 .0180968 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-35, the variances were homogeneous. As the variances were homogeneous, the ANOVAs assumption of homogeneity of variances was not violated and therefore ANOVA test was considered to be appropriate and T-test was not used. Table 3-36. Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of construction productivity Levene Statistic df1 df2 Sig. Cost_growth .085 1 47 .772 Schedule_growth 1.100 1 46 .300

PAGE 63

63 Table 3-37. ANOVA table for cost growth and schedule growth when analyzing the impact of construction productivity Sum of Squares df Mean Square F Sig. Cost_growth Between Groups .026 1 .026 2.755 .104 Within Groups .446 47 .009 Total .473 48 Schedule_grow th Between Groups .007 1 .007 .713 .403 Within Groups .453 46 .010 Total .460 47 Box plots for cost growth and schedule growth when analyzing the impact of construction productivity can be seen in Figure 3-19. A B Figure 3-19. Box plot when analyzing the impact of construction productivity. A) Cost growth. B) Schedule growth. Mean growth comparison for impact of construction productivity is seen in Figure 3-20. A B Figure 3-20. Mean growth comparison when analyzing the impact of construction productivity. A) Cost growth. B) Schedule growth.

PAGE 64

64 Conclusion: Mean project cost growth for highly positive impact of construction productivity was not statistically significantly different than mean project cost growth for less than highly positive impact of construction productivity (p=0.104). However, mean project cost growth factor for highly positive impact of construction productivity was found to be -5.9% and mean project cost growth factor for less than highly positive impact of construction productivity was found to be -1%. Mean project schedule growth for highly positive impact of construction productivity was not statistically significantly different than mean project schedule growth for less than highly positive impact of construction productivity (p=0.403). However, mean project schedule growth factor for highly positive impact of construction productivity was found to be +0.9% and mean project schedule growth factor for less than highly positive impact of construction productivity was found to be +3.4%. Project team communication: Project team communication involves development of an overall communication plan, identifying methods of cutting down cultural and language barriers in order to facilitate free flow of ideas and information, identifying networks and responsibilities to facilitate proper dissemination of information, introducing informal communication to meetings, structuring meetings and objectives and building relationships among all project team participants, in order to ensure overall success of the project. The survey question in the questionnaire was stated as, using a scale from -5 to +5, where -5 means an extremely negative impact compared to what was expected or planned and +5 means an extremely positive impact compared to what was expected or planned, please indicate the extent to which project team communication had a net positive impact, a net negative impact, or was essentially as planned?

PAGE 65

65 Table 3-38. CII BM&M survey question for project team communication Extremely Negative Impact As Planned Extremely Positive Impact -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 The following hypothesis was used for the analysis in terms of cost growth: Ho: There is no difference between mean project cost growth when the project team communication has highly positive impact and when project team communication has less than highly positive impact. Ha: There is a significant difference between mean project cost growth when the project team communication has highly positive impact and when project team communication has less than highly positive impact. The following hypothesis was used for the analysis in terms of schedule growth: Ho: There is no difference between mean project schedule growth when the project team communication has highly positive impact and when project team communication has less than highly positive impact. Ha: There is a significant difference between mean project schedule growth when the project team communication has highly positive impact and when project team communication has less than highly positive impact. For analysis, scale of +1 to +5 was considered to be as highly positive impact and scale of 5 to 0 was considered to be less than highly positive impact. Data was combined in two groups. Highly positive impact was considered group 1 and less than highly positive impact was considered group 2. QQ plots to check normality of data can be seen in Figure 3-21. A B Figure 3-21. QQ plot when analyzing the impact of project team communication. A) Cost growth. B) Schedule growth.

PAGE 66

66 One way ANOVA analysis was performed in order to compare the means of project cost growth and schedule growth for highly positive impact and less than highly positive impact. Descriptive statistics and ANOVA table can be seen in Table 3-38 and Table 3-40. Table 3-39. Descriptive statistics for cost growth and schedule growth when analyzing the impact of project team communication N Mean Cost_growth 1.00000 26 -.0716906 2.00000 24 .0072365 Total 50 -.0338056 Schedule_growth 1.00000 25 -.0235760 2.00000 24 .1144365 Total 49 .0440220 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-39, the variances were homogeneous. Table 3-40. Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of project team communication Levene Statistic df1 df2 Sig. Cost_growth 1.117 1 48 .296 Schedule_growth 1.084 1 47 .303 Table 3-41. ANOVA table for cost growth and schedule growth when analyzing the impact of project team communication Sum of Squares df Mean Square F Sig. Cost_growth Between Groups .078 1 .078 6.549 .014 Within Groups .570 48 .012 Total .648 49 Schedule_grow th Between Groups .233 1 .233 6.064 .018 Within Groups 1.808 47 .038 Total 2.041 48 Box plots for cost growth and schedule growth when analyzing the impact of project team communication can be seen in Figure 3-22.

PAGE 67

67 A B Figure 3-22. Box plot when analyzing the impact of project team communication. A) Cost growth. B) Schedule growth. Mean cost growth and schedule growth comparison when analyzing the impact of project team communication can be seen in Figure 3-23. A B Figure 3-23. Mean growth comparison when analyzing the impact of project team communication. A) Cost growth. B) Schedule growth. Conclusion: Mean project cost growth for highly positive impact of project team communication was statistically significantly different than mean project cost growth for less than highly positive impact of project team communication at 95% significance level (p=0.014). Mean project cost growth factor for highly positive impact of project team communication was found to be -7.2% and mean project cost growth factor for less than highly positive impact of project team communication was found to be +0.7%.

PAGE 68

68 Mean project schedule growth for highly positive impact of project team communication was statistically significantly different than mean project schedule growth for less than highly positive impact of project team communication at 95% significance level (p=0.018). Mean project schedule growth factor for highly positive impact of project team communication was found to be -2.4% and mean project schedule growth factor for less than highly positive impact of project team communication was found to be +11.4%. Project team expertise: Project team expertise involves experience levels of project team members such as project manager, safety manager, superintendent, contractors and subcontractors. The survey question in the questionnaire was stated as, using a scale from -5 to +5, where -5 means an extremely negative impact compared to what was expected or planned and +5 means an extremely positive impact compared to what was expected or planned, please indicate the extent to which project team expertise had a net positive impact, a net negative impact, or was essentially as planned? Table 3-42. CII BM&M survey question for project team expertise Extremely Negative Impact As Planned Extremely Positive Impact -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 The following hypothesis was used for the analysis in terms of cost growth: Ho: There is no difference between mean project cost growth when the project team expertise has highly positive impact and when project team expertise has less than highly positive impact. Ha: There is a significant difference between mean project cost growth when the project team expertise has highly positive impact and when project team expertise has less than highly positive impact.

PAGE 69

69 The following hypothesis was used for the analysis in terms of schedule growth: Ho: There is no difference between mean project schedule growth when the project team expertise has highly positive impact and when project team expertise has less than highly positive impact. Ha: There is a significant difference between mean project schedule growth when the project team expertise has highly positive impact and when project team expertise has less than highly positive impact. For analysis, scale of +1 to +5 was considered to be as highly positive impact and scale of 5 to 0 was considered to be less than highly positive impact. Data was combined in two groups. Highly positive impact was considered group 1 and less than highly positive impact was considered group 2. QQ plots to check normality of data can be seen in Figure 3-24. A B Figure 3-24. QQ plot when analyzing the impact of project team expertise. A) Cost growth. B) Schedule growth. One way ANOVA analysis was performed in order to compare the means of project cost growth and schedule growth for highly positive impact and less than highly positive impact. Descriptive statistics and ANOVA table can be seen in Table 3-42 and Table 3-44. Table 3-43. Descriptive statistics for cost growth and schedule growth when analyzing the impact of project team expertise N Mean Cost_growth 1.00000 29 -.0561228 2.00000 19 -.0398571 Total 48 -.0496843 Schedule_growth 1.00000 30 -.0029296 2.00000 19 .1181560 Total 49 .0440220

PAGE 70

70 Levenes statistic was used to check the homogeneity of variances. As seen in Table 3-43, the variances were homogeneous. Table 3-44. Test of homogeneity of variances for cost growth and schedule growth when analyzing the impact of project team expertise Levene Statistic df1 df2 Sig. Cost_growth .096 1 46 .758 Schedule_growth 2.175 1 47 .147 Table 3-45. ANOVA table for cost growth and schedule growth when analyzing the impact of project team expertise Sum of Squares df Mean Square F Sig. Cost_growth Between Groups .003 1 .003 .411 .525 Within Groups .340 46 .007 Total .343 47 Schedule_grow th Between Groups .171 1 .171 4.285 .044 Within Groups 1.871 47 .040 Total 2.041 48 Box plots for cost growth and schedule growth when analyzing the impact of project team expertise can be seen in Figure 3-25. A B Figure 3-25. Box plot when analyzing the impact of project team expertise. A) Cost growth. B) Schedule growth. Mean cost growth and schedule growth comparison when analyzing the impact of project team expertise can be seen in Figure 3-26.

PAGE 71

71 A B Figure 3-26. Mean growth comparison when analyzing the impact of project team expertise. A) Cost growth. B) Schedule growth. Conclusion: Mean project cost growth for highly positive impact of project team expertise was not statistically significantly different than mean project cost growth for less than highly positive impact of project team expertise (p=0.525). However, mean project cost growth factor for highly positive impact of project team expertise was found to be -8.9% and mean project cost growth factor for less than highly positive impact of project team expertise was found to be -8%. Mean project schedule growth for highly positive impact of project team expertise was statistically significantly different than mean project schedule growth for less than highly positive impact of project team expertise at 95% significance level (p=0.044). Mean project schedule growth factor for highly positive impact of project team expertise was found to be -4% and mean project schedule growth factor for less than highly positive impact of project team expertise was found to be -2.4%. 3.6.2 Regression Analysis Initially, a simple linear regression analysis was performed without any data grouping to see if there was any correlation of factors affecting construction quality with project cost growth and project schedule growth. Table 3-45 shows the summary of the results.

PAGE 72

72 Table 3-46. Summary of regression analysis Factor R 2 for Cost Growth Analysis P-value R 2 for Schedule Growth Analysis P-value Alliance activeness 0.34 0.000062 0.124 0.027 Alliance duration 0.04 0.218 0.0006 0.88 Availability of skilled labor 0.0009 0.84 0.0017 0.76 Materials availability/cost 0.10 0.02 0.01 0.48 Project team expertise 0.0066 0.57 0.04 0.16 Project team communication 0.0733 0.06 0.02 0.323 Construction productivity 0.0078 0.54 0.0003 0.9 Rework cost 0.04 0.004 0.0007 0.79 Each factor was treated as an independent variable and cost growth and schedule growth were treated as dependent variables. Regression analysis was performed one by one between each factor and project cost growth and project schedule growth separately. The hypothesis used for the analysis for each factor is as follows: Ho: Bi = 0 (i= 1 to 8) (all model terms are unimportant in predicting project cost growth) Ha: At-least one value of Bi 0 (at-least one model term is useful in predicting project cost growth) Ho: Bi = 0 (i= 1 to 8) (all model terms are unimportant in predicting project schedule growth) Ha: At-least one value of Bi 0 (at-least one model term is useful in predicting project schedule growth) As it can been clearly seen, the R2 values for all the factors are very low, the analysis did not yield any significant correlation. The detailed regression outputs from microsoft excel can be seen in the Appendix A.

PAGE 73

73 3.6.3 Curve Estimation Analysis As the simple linear regression analysis did not provide significant results, a curve estimation analysis was performed using SPSS Statistics 17.0, to find out if there was a possibility of non linear relationship between factors affecting construction quality and project cost growth and project schedule growth. In order to this average project cost growth and project schedule growth values were used instead of absolute values as used earlier in simple linear regression. For instance, scale of 1 7 signifying essentially an alliance on paper a very active alliance was used to collect data for alliance activeness between contractor and owner. So, average value of project cost growth and project schedule growth were calculated for each point on this 7 point scale. Data was prepared in the following way, Table 3-46. Table 3-47. Data preparation example for curve estimation analysis Alliance Activeness Rating Project Cost Growth Project Schedule Growth 1-7 Average value for cost growth for all the responses that considered 1-7 as the rating Average value for schedule growth for all the responses that considered 1-7 as the rating After this, the average values which seemed to be inconsistent with the rating value were omitted and were not used in the analysis. The curve estimation analysis provided the following results that are summarized in the Table 3-47. The cells marked in orange show results at 95% significance level and the cells marked in blue show results at 90% significance level. The table provides coefficient of correlations for possible linear, quadratic and cubic relationships of the eight factors with project cost growth and project schedule growth. It was observed that there were possible statistically significant non linear correlations of factors affecting quality of construction projects with project cost growth and project schedule growth. Detailed analysis tables can be seen in the Appendix B.

PAGE 74

74 Table 3-48. Curve estimation analysis summary Project Cost Growth Linear Quadratic Cubic Project Schedule Growth Linear Quadratic Cubic Alliance activeness R 2 = 0.868, p = 0.007 R 2 = 0.951, p = 0.011 R 2 = 0.955, p = 0.067 R 2 = 0.849, p = 0.079 R 2 = 0.948, p = 0.227 R 2 = 1.000, p = NA Alliance duration R 2 = 0.652, p = 0.193 R 2 = 0.848, p = 0.390 R 2 = 1.000, p = NA R 2 = 0.503, p = 0.291 R 2 = 0.999, p = 0.024 R 2 = 1.000, p = NA Skilled labor R 2 = 0.748, p = 0.058 R 2 = 0.945, p = 0.055 R 2 = 0.995, p = 0.093 R 2 = 0.985, p = 0.001 R 2 = 0.986, p = 0.014 R 2 = 0.999, p = 0.048 Materials availability R 2 = 0.388, p = 0.135 R 2 = 0.610, p = 0.152 R 2 = 0.852, p = 0.092 R 2 = 0.820, p = 0.034 R 2 = 0.828, p = 0.172 R 2 = 0.830, p = 0.510 Team expertise R 2 = 0.872, p = 0.020 R 2 = 0.882, p = 0.118 R 2 = 0.882, p = 0.118 R 2 = 0.768, p = 0.051 R 2 = 0.904, p = 0.096 R 2 = 0.919, p = 0.081 Team communication R 2 = 0.775, p = 0.021 R 2 = 0.836, p = 0.066 R 2 = 0.986, p = 0.021 R 2 = 0.503, p = 0.180 R 2 = 0.944, p = 0.056 R 2 = 0.983, p = 0.167 Construction productivity R 2 = 0.814, p = 0.014 R 2 = 0.921, p = 0.022 R 2 = 0.927, p = 0.108 R 2 = 0.710, p = 0.158 R 2 = 0.917, p = 0.287 R 2 = 0.917, p = 0.287 Rework cost R 2 = 0.846, p = 0.001 R 2 = 0.931, p = 0.001 R 2 = 0.974, p = 0.001 R 2 = 0.251, p = 0.097 R 2 = 0.258, p = 0.261 R 2 = 0.363, p = 0.282 After obtaining the statistical parameters for different types of possible relationships i.e. linear, quadratic and cubic, curves were generated to visualize these possible relationships between the factors and project cost growth and project schedule growth. These curves were generated using SPSS Statistics 17.0 software. Several iterations were performed step by step to remove the outliers in order to obtain the best-fit curves with high degree of correlation and high significance level. The following graphs in Figures 27-34 illustrate possibilities of non linear relationships between factors affecting construction quality and project cost growth, project schedule growth.

PAGE 75

75 A B Figure 3-27. Relationship of alliance activeness between contractors and owners with growth factors. A) Cost growth. B) Schedule growth. A B Figure 3-28. Relationship of alliance duration between contractors and owners with growth factors. A) Cost growth. B) Schedule growth. A B Figure 3-29. Relationship of availability of skilled labor with growth factors. A) Cost growth. B) Schedule growth.

PAGE 76

76 A B Figure 3-30. Relationship of materials availability/cost with growth factors. A) Cost growth. B) Schedule growth. A B Figure 3-31. Relationship of project team expertise with growth factors. A) Cost growth. B) Schedule growth. A B Figure 3-32. Relationship of project team communication with growth factors. A) Cost growth. B) Schedule growth.

PAGE 77

77 A B Figure 3-33. Relationship of construction productivity with growth factors. A) Cost growth. B) Schedule growth. A B Figure 3-34. Relationship of rework cost as % of actual project cost with growth factors. A) Cost growth. B) Schedule growth. 3.6.4 Multiple Linear Regression A multiple linear regression was performed taking project cost growth as the dependent variable and the factors that showed significant results in the ANOVA analysis as the independent variables. These factors were activeness of alliance relationship between contractor and owner, availability of skilled labor, materials availability/cost, project team communication and rework cost as % of actual project cost. The analysis gave a high correlation coefficient (R2 = 0.899). However, this correlation is not significant (p = 0.513). As found from the curve estimation discussed above, that there is a possibility of existence of non linear relationships

PAGE 78

78 between the variables and it was found that activeness of alliance relationship between contractor and owner vs. project cost growth and availability of skilled labor vs. project cost growth are better explained by cubic relationship. But, this multiple linear regression model assumes linear relationship between the factors and thus the significance level is low. The summary for the multiple regression parameters can be seen in Tables 3-48 3-50. Table 3-49. Model summary for multiple linear regression Model R R Square 1 .948 a .899 Table 3-50. ANOVA table for the multiple linear regression model Model Sum of Squares df Mean Square F Sig. 1 Regression .121 5 .024 1.779 .513 a Residual .014 1 .014 Total .135 6 Table 3-51. Correlation coefficients for the factors in multiple linear regression model Model Unstandardized Coefficients B Std. Error Standardized Coefficients Beta t Sig. 1 (Constant) .916 1.418 .646 .635 Allyactive -.061 .092 -.386 -.661 .628 Labor .028 .106 .091 .264 .836 Materials -.076 .070 -.775 -1.093 .472 Teamcomm -.056 .052 -.572 -1.083 .475 Rework .021 .065 .195 .325 .800 3.6.5 Multiple Linear Regression Analysis to Develop Scorecard The curve estimation analysis showed that there was possible significant linear relationship for following factors with project cost growth: Alliance activeness Skilled labor Team expertise Team communication Construction productivity Rework cost

PAGE 79

79 A multiple linear regression was run on these factors and associated project cost growth values. After the first iteration it was found that alliance activeness, team expertise and construction productivity hindered the degree of correlation and significance of the model. Therefore these factors were removed from the model step by step and a final model was obtained with skilled labor, team communication and rework cost as the independent variables and it was statistically significant at 95% significance level (p = 0.000) with R2 value = 0.936. The curve estimation analysis showed that there was possible significant linear relationship for following factors with project schedule growth: Alliance activeness Skilled labor Materials availability Team expertise Rework cost A multiple linear regression was run on these factors and associated project schedule growth values. Alliance activeness and rework cost were the factors which were found to have negative effect on the degree of correlation and significance of the model. Therefore, these two factors were removed from the model step by step and the regression analysis was performed again with the remaining variables i.e. availability of skilled labor, materials availability and project team expertise. A final model obtained with skilled labor, team expertise and materials availability as the independent variables was statistically significant at 95% significance level (p = 0.000) with R2 value = 0.910. The analysis results can be seen in Tables 3-51 3-56. Table 3-52. Model summary for multiple regression analysis for predicting project cost growth Model R R Square Adjusted R Square Std. Error of the Estimate 1 .967 a .936 .908 .01979144

PAGE 80

80 Table 3-53. ANOVA table for multiple regression analysis for predicting project cost growth Model Sum of Squares df Mean Square F Sig. 1 Regression .040 3 .013 33.854 .000 a Residual .003 7 .000 Total .043 10 Table 3-54. Correlations for multiple regression analysis for predicting project cost growth Model Unstandardized Coefficients B Std. Error Standardized Coefficients Beta t Sig. 1 (Constant) .360 .062 5.777 .001 Labor -.052 .006 -1.090 -9.339 .000 Team_comm -.026 .006 -.540 -4.538 .003 Rework .008 .002 .332 3.375 .012 Table 3-55. Model summary for multiple regression analysis for predicting project schedule growth Model R R Square Adjusted R Square Std. Error of the Estimate 1 .954 a .910 .880 .01250522 Table 3-56. ANOVA table for multiple regression analysis for predicting project schedule growth Model Sum of Squares df Mean Square F Sig. 1 Regression .014 3 .005 30.452 .000 a Residual .001 9 .000 Total .016 12 Table 3-57. Correlations for multiple regression analysis for predicting project schedule growth Model Unstandardized Coefficients B Std. Error Standardized Coefficients Beta t Sig. 1 (Constant) .365 .035 10.588 .000 Labor -.024 .003 -.824 -7.659 .000 Materials -.014 .003 -.455 -4.244 .002 Team_exp -.016 .002 -.819 -7.193 .000 The scorecard asks the respondents to evaluate factors impacting project cost growth and project schedule growth that were found to be significant through regression analysis done for developing the scorecard explained above. Based on the score entered for these factors equations

PAGE 81

81 for estimated cost growth and estimated schedule growth can be used. The following Figure 3-57 demonstrates the scorecard. Figure 3-35. Scorecard for estimating project cost growth and project schedule growth btbnfnrbnbnbbbn bbbb b!b "b#$b bb%b&!'b!(b)!*b)+b!b#$!b#b&)!+bb)+b+b bb b %&!'b)!(b)!*b)+b!b#$!b#b&)!+bb)+"b)b+!b!$b&!!b!b#$$b$bb!$b#b!b $bb!b)!(b)!"bb!b!(b)!"bbb!'bb)+,b &!'b !( b )+ &!'b )!( t!b bbbbbb / 0b(1!'bb2+b1b-/ 3456789: ; <0b=>!b!b?!b-n/ 3456789: ; 0b#2b!bbb@bb!?b)>!b !b-/ !!+b!b#!$b-000/ 000bAb;047;b.;0;63-/b.;0;37-n/bBb;0;;9-/bAb 0b(1!'bb2+b1b-/ 3456789: ; <0bf!b(1!'b-f/ 3456789: ; 0b=>!b!b&)!b-n/ 3456789: ; !!+b$+?b#!$b-000/ 000bAb;0476b.;0;35-/b.;0; 5-f/b.;0; 7-n/bAb

PAGE 82

82 CHAPTER 4 CONCLUSIONS AND RECOMMENDATIONS 4.1 Conclusions The conclusions for the analysis for the eight factors affecting quality of construction projects are discussed in this chapter. Based on the factors that were identified through extensive literature review and the questions from the Construction Industry Institute (CII) Benchmarking & Metrics (BM&M) questionnaire that addressed these factors, the following eight factors were identified and analyzed to determine their relationship with project cost growth and project schedule growth: Activeness of alliance between contractors and owners Duration of alliance relationship between contractors and owners Availability of skilled labor Materials availability/cost Project team expertise Project team communication Construction productivity Rework cost as a % of actual project cost From the statistical analysis it was found that project cost growth was impacted by all the eight factors but significantly by following five factors: Activeness of alliance between contractors and owners Availability of skilled labor Materials availability/cost Project team communication Rework cost as a % of actual project cost Possible reasons for non-significant results for the remaining three factors i.e., duration of alliance relationship between contractors and owners, project team expertise, construction productivity are explained: Duration of alliance relationship between contractors and owners: there exists a possibility that project cost growth is not always affected positively by number of years of alliance between contractor and owner. It may depend on commitment towards work

PAGE 83

83 and on activeness of alliance and not on just number of years. A contractor working for the first time is also capable of providing a good project outcome. Project team expertise: although project cost growth comparison between highly positive impact of project team expertise and less than highly positive impact of project team expertise was not statistically significant, but the curve estimation analysis showed that there was a significant correlation between project cost growth and project team expertise. Construction productivity: although project cost growth comparison between highly positive impact of construction productivity and less than highly positive impact of construction productivity was not statistically significant, but the curve estimation analysis showed that there was a significant correlation between project cost growth and construction productivity. It was also concluded that the lower project schedule growth was associated with high positive impact of all the eight factors but the significant result was found only for project team expertise and project team communication. However, the curve estimation analysis showed that there was a significant correlation of all the eight factors with project schedule growth and indicated that lower project schedule growth was associated with high positive impact of all the eight factors but as an exception, the curve estimation analysis showed that project schedule growth decreases with increase in rework cost. This conclusion seems to be unlikely in reality. Time required to do the rework would have been a better parameter to compare with project schedule growth rather than cost. The non significant results for analysis involving testing of the project schedule growths association with the eight factors can be explained by the fact that the respondent who must have filled the CII BM&M survey must have overlooked the schedule impacts of these factors during the time of project execution, while filling out the survey, and must have taken into account just the final schedule outcome.

PAGE 84

84 4.1.1 Possible Relationships of Factors with Project Cost Growth It was observed that the factors affecting quality of construction projects have other than a linear relationship with project cost growth while some were explained better by linear relationships. Alliance activeness between contractor and owner is better explained by a cubic relationship. Alliance duration between contractor and owner is better explained by a cubic relationship. Availability of skilled labor is better explained by a quadratic relationship. Materials availability/cost is better explained by a cubic relationship. Project team expertise is better explained by a linear relationship. Project team communication is better explained by a quadratic relationship. Construction productivity is better explained by a linear relationship. Rework cost is better explained by a quadratic relationship. 4.1.2 Possible Relationships of Factors with Project Schedule Growth It was observed that the factors affecting quality of construction projects have other than a linear relationship with project schedule growth while some were explained better by linear relationships. Alliance activeness between contractor and owner is better explained by a linear relationship. Alliance duration between contractor and owner is better explained by a quadratic relationship. Availability of skilled labor is better explained by linear relationship. Materials availability/cost is better explained by a linear relationship. Project team expertise is better explained by a cubic relationship. Project team communication is better explained by a quadratic relationship.

PAGE 85

85 Construction productivity could not be explained in terms of significant linear, quadratic or cubic relationship. Rework cost is better explained by a linear relationship but this relationship is unlikely to happen in reality. This may be due to the fact that the number of days required to do rework may have been a better factor to compare with schedule growth. The multiple linear regression analysis was performed between significant factors and project cost growth yielded a non-significant model because of assumption of linear relationships with all the factors. 4.1.3 Scorecard for Estimating Project Cost Growth and Project Schedule Growth The developed scorecard can be used for the estimation of project cost growth assuming linear relationship of cost growth with availability of skilled labor, project team communication and rework cost as a % of actual project cost and can be used for the estimation of project schedule growth assuming linear relationship of schedule growth with availability of skilled labor, project team expertise and materials availability/cost. The project manager can evaluate the factors listed on the scorecard and give them a score based on actual conditions on jobsite. The scorecard provides two equations for estimated cost growth and estimated schedule growth respectively. These equations use the scores decided by the project manager for different factors and provide the estimated value of cost growth and schedule growth. Therefore, the use of scorecard will help the project managers to assess the current situation on job site in terms of desired project outcome and thus will help them to take appropriate actions to rectify the problems proactively. 4.2 Limitation of Research The research used the data that had been collected by Construction Industry Institute (CII) using there own benchmarking questionnaire. Data was not available to address all the factors according to there possible metrics as identified through literature review.

PAGE 86

86 4.3 Recommendations This research analyzed construction project quality in terms of project cost growth and project schedule growth. Further studies on the cost impact, schedule impact and customer satisfaction of various factors affecting construction quality are recommended. Although this research analyzed relationship of factors affecting construction quality with both project cost growth and project schedule growth individually, the analysis should be expanded to analyze the data for multiple regression, so that the integrated impact caused by these factors can be identified. It was also seen that there existed a possibility of non linear relationships of factors affecting construction quality with both project cost growth and project schedule growth. Therefore, studies on the establishing the accurate non linear relationship for individual relationships between factors affecting construction quality and project cost growth, project schedule growth and customer satisfaction and for the integrated model that combines the performance variables i.e. cost, schedule and customer satisfaction and analyzes the impact of factors affecting quality on these variables using multiple regression is recommended and this would help in more accurate real time forecasting. As a lot of inconsistencies in the data were spotted, a final recommendation for future study is to develop a more accurate system for data collection at the project level which will help to establish more accurate relationships.

PAGE 87

87 APPENDIX A SUMMARY OF REGRESSION ANALYSIS FOR CURVE ESTIMATION Table A-1. Summary of regression analysis for curve estimation between cost growth and alliance activeness Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .868 26.366 1 4 .007 .343 -.064 Quadratic .951 28.896 2 3 .011 .577 -.185 .013 Cubic .955 14.019 3 2 .067 .730 -.310 .044 -.002 Table A-2. Summary of regression analysis for curve estimation between schedule growth and alliance activeness Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .849 11.244 1 2 .079 .280 -.035 Quadratic .948 9.192 2 1 .227 .136 .044 -.009 Cubic 1.000 3 0 -.215 .342 -.082 .005 Table A-3. Summary of regression analysis for curve estimation between cost growth and alliance duration Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .652 3.745 1 2 .193 -.210 .064 Quadratic .848 2.786 2 1 .390 -.405 .258 -.039 Cubic 1.000 3 0 .132 -.596 .345 -.051 Table A-4. Summary of regression analysis for curve estimation between schedule growth and alliance duration Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .503 2.023 1 2 .291 -.138 .075 Quadratic .999 877.854 2 1 .024 -.552 .489 -.083 Cubic 1.000 3 0 -.508 .419 -.051 -.004 Table A-5. Summary of regression analysis for curve estimation between cost growth and availability of skilled labor Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .748 8.909 1 3 .058 .062 -.022 Quadratic .945 17.110 2 2 .055 -.086 .057 -.009 Cubic .995 65.523 3 1 .093 .113 -.109 .033 -.003

PAGE 88

88 Table A-6. Summary of regression analysis for curve estimation between schedule growth and availability of skilled labor Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .985 198.803 1 3 .001 .090 -.013 Quadratic .986 72.642 2 2 .014 .084 -.010 .000 Cubic .999 232.902 3 1 .048 .131 -.047 .008 .000 Table A-7. Summary of regression analysis for curve estimation between cost growth and materials availability Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .388 3.175 1 5 .135 .192 -.035 Quadratic .610 3.128 2 4 .152 .493 -.168 .012 Cubic .852 5.761 3 3 .092 1.323 -.757 .132 -.007 Table A-8. Summary of regression analysis for curve estimation between schedule growth and materials availability Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .820 13.690 1 3 .034 .085 -.004 Quadratic .828 4.801 2 2 .172 .081 -.002 .000 Cubic .830 1.625 3 1 .510 .087 -.007 .001 -5.997E-5 Table A-9. Summary of regression analysis for curve estimation between cost growth and project team expertise Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .872 20.493 1 3 .020 .044 -.013 Quadratic .882 7.474 2 2 .118 .089 -.025 .001 Cubic .882 7.474 2 2 .118 .089 -.025 .001 .000 Table A-10. Summary of regression analysis for curve estimation between schedule growth and project team expertise Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .768 9.915 1 3 .051 .063 -.008 Quadratic .904 9.396 2 2 .096 .146 -.034 .002 Cubic .919 11.286 2 2 .081 .120 -.022 .000 7.921E-5

PAGE 89

89 Table A-11. Summary of regression analysis for curve estimation between cost growth and project team communication Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .775 13.778 1 4 .021 .188 -.037 Quadratic .836 7.660 2 3 .066 -.007 .041 -.007 Cubic .986 47.943 3 2 .021 -1.182 .768 -.147 .008 Table A-12. Summary of regression analysis for curve estimation between schedule growth and project team communication Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .503 3.040 1 3 .180 .052 -.012 Quadratic .944 16.723 2 2 .056 -.156 .065 -.006 Cubic .983 18.837 3 1 .167 .102 -.069 .014 .000 Table A-13. Summary of regression analysis for curve estimation between cost growth and construction productivity Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .814 17.494 1 4 .014 .092 -.024 Quadratic .921 17.391 2 3 .022 .238 -.083 .005 Cubic .927 8.412 3 2 .108 .398 -.179 .022 .000 Table A-14. Summary of regression analysis for curve estimation between schedule growth and construction productivity Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .710 4.892 1 2 .158 .374 -.047 Quadratic .917 5.560 2 1 .287 .796 -.206 .013 Cubic .917 5.560 2 1 .287 .796 -.206 .013 .000 Table A-15. Summary of regression analysis for curve estimation between cost growth and rework cost Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .846 32.860 1 6 .001 -.048 .024 Quadratic .931 33.941 2 5 .001 .049 -.004 .001 Cubic .974 49.424 3 4 .001 -.065 .070 -.009 .000

PAGE 90

90 Table A-16. Summary of regression analysis for curve estimation between schedule growth and rework cost Equation R Square F df1 df2 Sig. Constant b1 b2 b3 Linear .251 3.356 1 10 .097 .288 -.021 Quadratic .258 1.563 2 9 .261 .317 -.031 .001 Cubic .363 1.519 3 8 .282 .160 .079 -.015 .001

PAGE 91

91 APPENDIX B SUMMARY OF INITIAL REGRESSION ANALYSIS WITHOUT DATA GROUPING Table B-1. Summary output for regression between cost growth & alliance activeness Regression Statistics Multiple R 0.58356 R Square 0.34055 Adjusted R Square 0.32364 Standard Error 0.10190 Observations 41 df SS MS F Significance F Regression 1 0.20910 0.20910 20.13983 0.00006 Residual 39 0.40492 0.01038 Total 40 0.61403 Coefficients Standard Error t Stat P-value Intercept 0.22226 0.05384 4.12818 0.00019 Ally_active -0.04244 0.00946 -4.48774 0.00006 Table B-2. Summary output for regression between schedule growth & alliance activeness Regression Statistics Multiple R 0.35281 R Square 0.12447 Adjusted R Square 0.10081 Standard Error 0.16066 Observations 39 df SS MS F Significance F Regression 1 0.13577 0.13577 5.26023 0.02760 Residual 37 0.95499 0.02581 Total 38 1.09076 Coefficients Standard Error t Stat P-value Intercept 0.30044 0.09410 3.19280 0.00287 Ally_active -0.03766 0.01642 -2.29352 0.02760

PAGE 92

92 Table B-3. Summary output for regression between cost growth & alliance duration Regression Statistics Multiple R 0.20722 R Square 0.04294 Adjusted R Square 0.01560 Standard Error 0.12329 Observations 37 df SS MS F Significance F Regression 1 0.02387 0.02387 1.57040 0.21846 Residual 35 0.53198 0.01520 Total 36 0.55585 Coefficients Standard Error t Stat P-value Intercept -0.10998 0.08288 -1.32694 0.19312 Ally_duration 0.03005 0.02398 1.25315 0.21846 Table B-4. Summary output for regression between schedule growth & alliance duration Regression Statistics Multiple R 0.02439 R Square 0.00059 Adjusted R Square -0.02796 Standard Error 0.17573 Observations 37 df SS MS F Significance F Regression 1 0.00064 0.00064 0.02083 0.88607 Residual 35 1.08087 0.03088 Total 36 1.08151 Coefficients Standard Error t Stat P-value Intercept 0.07774 0.11391 0.68244 0.49945 Ally_duration 0.00478 0.03315 0.14433 0.88607

PAGE 93

93 Table B-5. Summary output for regression between cost growth & availability of skilled labor Regression Statistics Multiple R 0.03037 R Square 0.00092 Adjusted R Square -0.02033 Standard Error 0.11724 Observations 49 df SS MS F Significance F Regression 1 0.00060 0.00060 0.04339 0.83590 Residual 47 0.64598 0.01374 Total 48 0.64657 Coefficients Standard Error t Stat P-value Intercept -0.02112 0.06603 -0.31989 0.75047 Labor -0.00233 0.01118 -0.20830 0.83590 Table B-6. Summary output for regression between schedule growth & availability of skilled labor Regression Statistics Multiple R 0.04187 R Square 0.00175 Adjusted R Square -0.01949 Standard Error 0.24503 Observations 49 df SS MS F Significance F Regression 1 0.00496 0.00496 0.08254 0.77515 Residual 47 2.82176 0.06004 Total 48 2.82672 Coefficients Standard Error t Stat P-value Intercept 0.02364 0.13801 0.17126 0.86476 Labor 0.00671 0.02336 0.28730 0.77515

PAGE 94

94 Table B-7. Summary output for regression between cost growth and materials availability/cost Regression Statistics Multiple R 0.32175 R Square 0.10352 Adjusted R Square 0.08445 Standard Error 0.11105 Observations 49 df SS MS F Significance F Regression 1 0.06693 0.06693 5.42738 0.02417 Residual 47 0.57964 0.01233 Total 48 0.64657 Coefficients Standard Error t Stat P-value Intercept 0.14122 0.07705 1.83290 0.07316 Materials -0.03141 0.01348 -2.32967 0.02417 Table B-8. Summary output for regression between schedule growth & materials availability/cost Regression Statistics Multiple R 0.10445 R Square 0.01091 Adjusted R Square -0.01013 Standard Error 0.24390 Observations 49 df SS MS F Significance F Regression 1 0.03084 0.03084 0.51840 0.47509 Residual 47 2.79588 0.05949 Total 48 2.82672 Coefficients Standard Error t Stat P-value Intercept -0.05724 0.16922 -0.33824 0.73669 Materials 0.02132 0.02961 0.72000 0.47509

PAGE 95

95 Table B-9. Summary output for regression between cost growth & project team expertise Regression Statistics Multiple R 0.08158 R Square 0.00666 Adjusted R Square -0.01404 Standard Error 0.11576 Observations 50 df SS MS F Significance F Regression 1 0.00431 0.00431 0.32162 0.57328 Residual 48 0.64321 0.01340 Total 49 0.64752 Coefficients Standard Error t Stat P-value Intercept -0.07728 0.07839 -0.98587 0.32914 Team_experience 0.00591 0.01042 0.56711 0.57328 Table B-10. Summary output for regression between schedule growth & project team expertise Regression Statistics Multiple R 0.20384 R Square 0.04155 Adjusted R Square 0.02158 Standard Error 0.23758 Observations 50 df SS MS F Significance F Regression 1 0.11745 0.11745 2.08081 0.15565 Residual 48 2.70928 0.05644 Total 49 2.82673 Coefficients Standard Error t Stat P-value Intercept 0.28888 0.16088 1.79564 0.07885 Team_experience -0.03084 0.02138 -1.44250 0.15565

PAGE 96

96 Table B-11. Summary output for regression between cost growth & project team communication Regression Statistics Multiple R 0.27087 R Square 0.07337 Adjusted R Square 0.05407 Standard Error 0.11180 Observations 50 df SS MS F Significance F Regression 1 0.04751 0.04751 3.80073 0.05709 Residual 48 0.60001 0.01250 Total 49 0.64752 Coefficients Standard Error t Stat P-value Intercept 0.08253 0.06173 1.33692 0.18755 Team_communicate -0.01662 0.00853 -1.94955 0.05709 Table B-12. Summary output for regression between schedule growth & project team communication Regression Statistics Multiple R 0.14243 R Square 0.02029 Adjusted R Square -0.00012 Standard Error 0.24020 Observations 50 df SS MS F Significance F Regression 1 0.05735 0.05735 0.99393 0.32378 Residual 48 2.76938 0.05770 Total 49 2.82673 Coefficients Standard Error t Stat P-value Intercept 0.18974 0.13263 1.43064 0.15901 Team_communicate -0.01826 0.01831 -0.99696 0.32378

PAGE 97

97 Table B-13. Summary output for regression between cost growth & construction productivity Regression Statistics Multiple R 0.08840 R Square 0.00782 Adjusted R Square -0.01286 Standard Error 0.11569 Observations 50 df SS MS F Significance F Regression 1 0.00506 0.00506 0.37810 0.54153 Residual 48 0.64246 0.01338 Total 49 0.64752 Coefficients Standard Error t Stat P-value Intercept -0.07760 0.07308 -1.06189 0.29360 Cons_product 0.00752 0.01224 0.61489 0.54153 Table B-14. Summary output for regression between schedule growth & construction productivity Regression Statistics Multiple R 0.01793 R Square 0.00032 Adjusted R Square -0.02051 Standard Error 0.24263 Observations 50 df SS MS F Significance F Regression 1 0.00091 0.00091 0.01543 0.90167 Residual 48 2.82582 0.05887 Total 49 2.82673 Coefficients Standard Error t Stat P-value Intercept 0.04338 0.15326 0.28303 0.77837 Cons_product 0.00319 0.02566 0.12421 0.90167

PAGE 98

98 Table B-15. Summary output for regression between cost growth & rework cost Regression Statistics Multiple R 0.19243 R Square 0.03703 Adjusted R Square 0.03261 Standard Error 0.20042 Observations 220 df SS MS F Significance F Regression 1 0.33670 0.33670 8.38269 0.00417 Residual 218 8.75630 0.04017 Total 219 9.09300 Coefficients Standard Error t Stat P-value Intercept -0.00690 0.01630 -0.42298 0.67273 Rework/actual 0.00994 0.00343 2.89529 0.00417 Table B-16. Summary output for regression between schedule growth & rework cost Regression Statistics Multiple R 0.02588 R Square 0.00067 Adjusted R Square -0.00932 Standard Error 0.74547 Observations 102 df SS MS F Significance F Regression 1 0.03725 0.03725 0.06704 0.79623 Residual 100 55.57277 0.55573 Total 101 55.61003 Coefficients Standard Error t Stat P-value Intercept 0.21458 0.09103 2.35723 0.02036 Rework cost -0.00590 0.02277 -0.25892 0.79623

PAGE 99

99 LIST OF REFERENCES Abdul, R. H. (1993). Capturing the cost of quality failures in civil engineering. International Journal of Quality & Reliability Management, 10(3), 20. Ahmed, S. M., Azhar, S., and Forbes, L. H. China Association of Occupational Safety and Health. Akintoye, A., McIntosh, G., and Fitzgerald, E. (2000). An analysis of success factors and benefits of partnering in construction. European Journal of Purchasing & Supply Management, 6(3), 159. Baiden, B. K., Price A. D. F., and Dainty A. R. J. (2003). Looking beyond processes: Human factors in team integration. Brighton: ARCOM, 233. Baiden, B. K., Price, A. D. F., and Dainty, A. R. J. (2006). The extent of team integration within construction projects. International Journal of Project Management, 13-23. Bourn, J. (2000). Modernising Construction, National Audit Office, London. Bromilow, F. J. (1974). Measurement and scheduling of construction time and cost performance in the building industry. Charter Build. Bryde, J. D. (2003). Modelling project management performance. International Journal of Quality & Reliability Management, 20(2), 225. Chan, A. P. C. (2001). Time-cost relationship of public sector projects in Malaysia. International Journal of Project Management, 9, 223. Chauvel, A., and Andre, Y. (1985). Quality cost: Better prevent than cure. Quality Progress, 18 (9), 29. Construction Industry Development Agency and Masters Builders Australia (CIDA). (1995). Measuring up or muddling through: Best practice in the Australian non-residential construction industry, Sydney, Australia. Dale, B. G., and Plunkett, J. (1991). Quality costing, Chapman and Hall, London. Davis, K., Ledbetter, W. B., and Burati, J. L. (1989). Measuring design and construction quality costs. Journal of Construction Engineering and Management, 115, 389. Dept. of Industry, Science and Tourism (DIST). (1998). Building for Growth: a draft strategy for the building and construction industry, Commonwealth of Australia, Canberra, Australia. Eddie, W. L. Cheng, H. L., and Fox, P. (2007). Job Performance Dimensions for Improving Final Project Outcomes. Journal of Construction Engineering and Management, 133, 8. Egan J. (2002). Accelerating Change, Department of the Environment, Transport and the Regions, London.

PAGE 100

100 Egan, J. (1998). Rethinking Construction, Department of the Environment, Transport and the Regions, London. Evans, M. (2005). Overdue and over budget, over and over again. Project Management. Economist, 375(8430), 66. Evbuomwan, N. F. O., and Anumba C. J. (1998). An integrated framework for concurrent lifecycle design and construction. Adv. Eng. Software, 29(7-9), 587-97. Federele, M. O., and Chase, G. W. (1993). Applying total quality management to design and construction. Journal of Management in Engineering, 9-4. Fergusson, K. J. (1993). Impact of integration on industrial facility quality, Stanford Univ., Palo Alto, Calif. Flyvbjerg, B., Holm, M. K. S., and Buhl, S. L. (2003). What causes cost overrun in transport infrastructure projects. Transp Rev, 24(1), 3. Glassop, L. I. (2002). The organisational benefits of teams. Human Relations, 55(2), 225. Griffis, F. H., Hogan, D. B., and Li, W. (1995). An analysis of the impacts of using three dimensional computer models in the management of construction. Construction Industry Institute (CII), Austin, Texas, 106-11. Griffith, A. F., Gibson, G. E., Hamilton, M. R., Tortora, A. L., and Wilson, C. T. (1999). Project Success Index for Capital Facility Construction Projects. Journal of Performance of Constructed Facilities, 13, 1. Hayes, N. (2002). Managing teams: a strategy for success, Thomson Learning, London. Hendrickson, C., and Tung, A. (1998). Project Management for Construction. Horner, R. M., and Duff, A. R. (2001). More for Less: a Contractor's Guide for Improving Productivity in Construction, Westminster, London. Howell, I. (1996). The need for interoperability in the construction industry. Ireland, V. (2004). Improving staff performance. Construction Management & Economics, 22(2), 121. Iyer, K. C., and Jha, K. N. (2005). Factors affecting cost performance: evidence from Indian construction projects. International Journal of Project Management, 23, 283. Josephson, P. E., Larsson, B., and Heng, L. (2002). Illustrative Benchmarking Rework and Rework Costs in Swedish Construction Industry. Journal of Management in Engineering, 18, 2. Kaka, A., and Price, A. D. F. (1991). Relationship between value and duration of construction projects. Construction Management & Economics, 9, 383.

PAGE 101

101 Kaliba, C., Muya, M., and Mumba, K. (2008). Cost escalation and schedule delays in road construction projects in Zambia. International Journalof Project Management. Kamrul, A., and Gunawan, I. (2009). Analysis of cost and schedule performance of international development projects. International Journal of Project Management. Krizek, R. J., Lo, W., and Hadavi, A. (1996). Lessons learned from multiphase reconstruction project. Journal of Construction Engineering and Management, 122(1), 44. Kumaraswamy, M. M., and Chan, D. W. M. (1995). Determinants of construction duration. Construction Management & Economics, 13, 209. Latham, M. (1994). Constructing the team: joint review of procurement and contractual arrangements in the United Kingdom construction industry: final report, Department of the Environment, London. Love P. E. D., and Gunasekaran A. (1998). Concurrent engineering: a strategy for procuring construction projects. International Journal of Project Management, 16(6), 375. Love, P. E. D., and Li, H. (1999a). Knocking down walls and building bridgesOvercoming the problems associated with quality certification. Construction Management & Economics. Love, P. E. D., and Li, H. (1999b). Determining the causal structure of rework influences in construction. Construction Management & Economics, 17(4), 505. Love, P. E. D., Wyatt, A. D., and Mohamed, S. (1997b). Understanding Rework in Construction. International Conference on Construction Process Re-engineering, 269278. Low, S. P., and Yeo, H. K. C. (1998). A construction quality costs quantifying system for the building industry. International Journal of Quality & Reliability Management, 15(3), 329. Matta, F., and Ashkenas, R. (2003). Why good projects fail anyway. Harvard Bus. Rev., 81(9), 109. McKim, R., Hegazy, T., and Attalla, M. (2000). Project Performance Control in Reconstruction Projects. Journal of Construction Engineering and Management, 126, 2. McTague, B. (2002). Productivity Improvements on Alberta Major Construction Projects. Moore, D. R., and Dainty, A. R. J. (1999). Integrated project teams performance in managing unexpected change events. Team Perform Management, 5(7), 212. Moore, D. R., and Dainty, A. R. J. (2001). Intra-team boundaries as inhibitors of performance improvement in UK design and build projects: a call for change. Construct Management & Economics, 19(6), 559.

PAGE 102

102 Neese, T. A., and Ledbetter, W. B. (1991). Quality Performance Management in Engineering/Construction. Journal AACE Transactions. OConnor, J. T., and Yang, L. R. (2004). Project Performance versus Use of Technologies at Project and Phase Levels. Journal of Construction Engineering and Management, 130, 3. Payne, J., Thomas, K., Perkins, M., Parker, R., and Small, J. (2003). Working in an integrated team, Construction Productivity Network, London. Qureshi, T. M., Warraich, A. S., and Hijazi, S. T. (2009). Significance of project management performance assessment (PMPA) model. International Journal of Project Management, 27, 378. Raftery, J. (1994). Risk analysis in project management, E and FN Spon., London. Russell, J. S. (1991). Contractor Failure: Analysis. Journal of Performance of Constructed Facilities, 5, 3. Russell, J. S., and Jaselskis, E. J. (1992). Quantitative Study of Contractor Evaluation Programs And Their Impact. Journal of Construction Engineering and Management, 118, 3. Russell, J. S., and Jaselskis, E. J. (1997). Continuous Assessment of Project Performance. Journal of Construction Engineering and Management, 123, 1. Schleifer, T. C. (1987). A contractor's survival guide managing with confidence. Sugiharto, A., Hampson, K., and Mohamed, S. (1999). Investigation into the Relationship Between Rework and Site Supervision in High Rise Building Construction in Indonesia, Sydney. Taneja, Y.R. (1994). Quality Assurance for Building and Construction Industry. Journal of Construction Management, IX, III, 140-149. Thomas, S. R., Macken, C. L., and Lee, S. H. (2001). Impacts of design/ information technology on building and industrial projects. Construction Industry Institute (CII), Austin, Texas, BMM2001-10 Vincent, S., and Kirkpatrick, S.W. (1995). Integrating different views of integration. Integrated construction information, 53. Vyse, S. (2001). Fusion: a new approach to working, GlaxoWellcome, London. Walt, L., Zwikael, O., Henderson, K., and Anbari, F. (2009). Prediction of project outcome: The application of statistical methods to earned value management and earned schedule performance indexes. International Journal of Project Management, 27, 400. Wuellner, W. W. (1990). Project Performance Evaluation Checklist for Consulting Engineers. Journal of Management in Engineering, 6, 3.

PAGE 103

103 Yung, P., and Yip, B. (2009). Construction quality in China during transition: A review of literature and empirical examination. International Journal of Project Management.

PAGE 104

104 BIOGRAPHICAL SKETCH Ankit Bansal was born in 1984 in Sri Ganganagar, India. The eldest of two children, he grew up mostly in New Delhi, India, graduating from St. Georges School in 2003. His fascination for structures, construction design grew along with Indias advancement in this field and identified it as his career option. Born and brought up in a middle class family, that had a construction business, the interest towards this field grew from childhood when he used to accompany his father on construction sites. His determination led him to pursue engineering from one of the most reputed engineering institutes of India, Thapar University. He completed B.E. in civil engineering in 2007. His first tryst with the practical world of construction was during the senior year of his undergraduate education, as a part of six-months project semester training at Nagarjuna Constructions Ltd. (NCL), Gurgaon, India. He had practical experience of site work as well as learned various design aspects in staddpro and autocad. He worked on a prestigious project of Delhi Metro Rail Corporation (DMRC) office building complex. This training validated the choice of his interest and convinced him of the need for advanced education in building construction management. He pursued his M.S. in building construction from M.E. Rinker, Sr. School of Building Construction, University of Florida and graduated in 2009. During the course, he worked as a research assistant, assisting research team 254 (Quality Management) formed by the Construction Industry Institute (CII), a consortium of more than 100 leading owner, engineering-contractor, and supplier firms who have joined together to enhance the business effectiveness and sustainability of the capital facility life cycle. He assisted in conducting in-depth surveys of CII members, analyzing the results and identifying maturity levels, key drivers and best practices in implementing and improving quality management systems governing the development of major capital facilities. This research experience influenced him to do an independent research in the area of project level quality management

PAGE 105

105 and thus he decided to write his masters thesis on project level factors affecting quality of construction projects. After completing his M.S., Ankit decided to return back to India and work for the construction industry with an aim to make significant contributions to the society.