Evaluation and analysis of criteria and sub-criteria of a ... - Emerald

A constructionexcellence model23Figure 1.Evaluation of constructionexcellence model

ECAM15,124sub-criteria of the model were measured in the analytical part of a questionnairesurvey. The criteria and sub-criteria were measured for both their importance andactual performance within the company of the respondent. Third, the data collectedwas prepared for further analysis through coding, missing data analysis, and dataexploration. Fourth, the questionnaire was evaluated as a measuring instrument interms of reliability and validity. Fifth, the criteria and sub-criteria were evaluatedthrough their importance measurement, via factor analysis, and through theireffectiveness measurement. Finally, weights of the model criteria were computed on anempirical basis using factor regression coefficients. The steps of evaluation arediscussed in the sections to follow.The proposed construction excellence modelThe construction excellence model under evaluation was developed and described inBassioni et al. (2004a, 2005) and is illustrated in Figure 2. The model is divided intoenabling and results criteria. The enabling criteria include: leadership; customers andstakeholder focus; strategic management; information and analysis; people,partnerships, suppliers, physical resources, intellectual capital, risk, work culture,and process management. The results criteria include: internal stakeholder; project,external stakeholder; and organisational business results. The logic of the model startswith leadership as the main driver for change and improvement in organisations.Leadership should guide the focus on customers, people and other relevantstakeholders, which in turn should guide the development of strategic plans. Thestrategic plans are further detailed into functional or programmatic business plans thatare translated into processes for implementation. Once implemented on projects andthroughout the organisation, improved internal stakeholder results should start toappear (internal in terms of being under direct influence of management), such as thatof employees, suppliers and other internal stakeholders. This will reflect on projectresults that would further affect customer and external stakeholder results not underthe direct management influence of the organisation. The outcome is business resultsbeing achieved on the organisational level. Work culture is seen to be driven byleadership, and affecting other enabling criteria, whereas information and analysis isalso driven by leadership and supports all other criteria throughout the model. Thebi-directional relationships between information and analysis and other criteriaprovide forward and backward loops of information among criteria.Measurement of model criteria and sub-criteriaThe measurement of the model’s criteria and sub-criteria was conducted through aquestionnaire survey. This section discusses the design of the questionnaire showinghow the criteria and sub-criteria were measured, and the sampling approach of thesurvey.Questionnaire designThe questionnaire is divided into three questions, as can be seen in the Appendix(Figure A1). The first question aims to identify the importance of each criterion inimproving organisational business performance results. It uses a five-point Likert scale

A constructionexcellence model25Figure 2.The suggestedconstruction excellencemodel

effectiveness scores of their underlying sub-criteria, thus giving an indication of theactual performance of each criterion.Missing values can cause problems with some statistical techniques and could beresolved by either deletion or replacement methods (Hair et al., 1998). A missing valueanalysis was conducted on the Statistical Package for Social Sciences (SPSS) softwareto evaluate the magnitude of missing data before deciding how best to handle them instatistical techniques. A maximum of two missing values were found in less than 3 percent of variables, which indicates a low percentage of missing values. The mean of thevariable was used to replace missing data (Hair et al., 1998). One respondent did notanswer any of the “sub-criterion importance” questions, and was therefore discardedfrom the analysis, thus leaving the sample of sub-criteria importance variables being49 responses.A constructionexcellence model27Data explorationThe mean and standard deviation of each observed variable were computed in themissing value analysis. The normality of the distribution of each variable wasassessed, where the skewness and kurtosis values were computed. The assessment ofnormality was initially based on the values of Z skewness and Z kurtosis that werecalculated from the following equations, as expressed by Hair et al. (1998, p. 72), interms of the sample size N:Z skewness ¼ skewness qffiffiZ kurtosis ¼ kurtosis qffiffiffi6NThese values were compared to 2.58 to assess the significance of the normalityassumption at the 0.01 level (Hair et al., 1998, p. 73). If the variable was found to benon-normal, it was further assessed for the degree of non-normality according to thecriteria of Curran et al. (1996 in Byrne, 1998). Curran et al.’s criteria considereddistributions with skewness between 2 and 3 and kurtosis in the range of 7 to 21 to bemoderately non-normal, whereas skewness of above 3 and kurtosis of above 21 wereconsidered extremely non-normal. The reason for this assessment of degree ofnon-normality is that many statistical techniques are robust and can withholddeviations from the normality assumption (Stevens, 2002, p. 263). The only variablethat has shown extreme non-normality was the importance of leadership indetermining business performance (Question 1-a). It should be noted though that thevariables showing moderate or extreme deviation from normality are all negativelypeaked, which means they have higher concentrations towards their maximum values.Evaluation of measuring instrumentThe measurement of variables in a theoretical framework is an integral part ofresearch, and abstract concepts should be measured through physical/actualmeasurement of their operational definitions (Sekaran, 2003). This concept ofdeveloping a measuring instrument in quality management research has been used toconfirm the abstract constructs/criteria of Excellence Models (Wilson and Collier, 2000)and TQM frameworks (Ahire et al., 1996; Anderson et al., 1995; Black and Porter, 1996;24N

improve reliability, SPSS provides an expected value of alpha if a certain item were tobe deleted. Item (sub-criterion) “d. Society and environmental impact of projects” wasfound to have the largest impact on reliability improvement and was, thus, deleted toimprove the reliability of the measuring instrument. The resulting Cronbach’s alphabecame 0.6406 as indicated in Table I. The deletion of this item is acceptable from acomputational standpoint since five items were used to express the underlyingconstruct, and the deletion of one would leave four items to express the criterion.However, in the case of the construct “organisational business performance”,Cronbach’s alpha was found to be 0.5702, but this criterion was defined by only twoitems and deletion of an item would essentially separate the concept between financialand non-financial performance. Moreover, the alpha value is quite close to theacceptable value of 0.6. Accordingly, and in order to represent the concept in itsentirety, both items are retained.A constructionexcellence model29Validity of measuring instrumentAs previously discussed, validity indicates that the instrument is measuring what it issupposed to measure. Three types of validity are most common in business andorganisational research: content validity; construct validity and criterion-relatedvalidity (Brewerton and Millward, 2001; Sekaran, 2003).Content validity. Defined as “the extent to which it provides adequate coverage ofthe topic under study” (Cooper and Emory, 1995, p. 149). This type of validity ismostly based on the analysis of the target domain required, and drawn on expertjudgement (Brewerton and Millward, 2001). Content validity is demonstrated in thisinstrument in two ways. First, the “analysis of the target domain” was achievedthrough the literature review conducted in previous papers, and the theoreticaldevelopment of the model based on well-established models (Bassioni et al., 2004a,b). Second, the “expert judgement” was based on the empirical evaluation of expertinterviews and case studies (Bassioni et al., 2005), and the evaluation of thequestionnaire in the pilot study.Construct validity. Shows the extent to which items of a construct measure the sameconstruct (Flynn et al., 1994), i.e. do not measure multiple constructs. Construct validitycan be demonstrated by a factor analysis on the items of each construct (Cooper andEmory, 1995; Flynn et al., 1994). Table II and Figure 3 show the factor analysis andassociated Scree plot conducted on the sub-criteria of leadership using SPSS software.ComponentInitialEigenvaluestotalPer centofvarianceCumulative%Extraction sums ofsquared loadingstotalPer centofvarianceCumulative%1 3.138 52.298 52.298 3.138 52.298 52.2982 0.964 16.071 68.3693 0.728 12.141 80.5104 0.542 9.027 89.5375 0.361 6.015 95.5526 0.267 4.448 100.000Table II.Factor analysis ofleadership sub-criteria

Importance of criteria in determining business performanceRespondents of the survey were asked to rate the importance of each criterion indetermining organisational business performance in Question 1 of the questionnaire.The results of this question are summarised in terms of the mean and lower 5 per centconfidence limit of each criterion, and illustrated in Figure 4. These results werecomputed on an Excel spreadsheet using its embedded functions. The line chart showsan above average rating of importance for all criteria and close to maximum rating formany of them. Leadership, customer focus and people management showed the highestrelevance, among all criteria, to improving organisational business performance. Theonly criterion having extreme non-normality (leadership), as discussed in the dataexploration sub-section was negatively peaked. Its computed confidence limit based onnormality is expected to provide an understatement of the actual lower confidence limits.A constructionexcellence model31Factor analysis of excellence criteriaConfirmatory factor analysis differs than exploratory factor analysis in the postulationof a factor structure, whereas in exploratory factor analysis the factor structure isdetermined within the analysis (Cramer, 2003). To ensure that the items load on onlyone variable a preliminary “exploratory factor analysis” is conducted on SPSS, then aconfirmatory factor analysis (CFA) is conducted using a specialised statistical softwareLISREL (LInear Structural RELationships).In the exploratory factor analysis, the method of principle component analysis wasused on SPSS and the scree plot of eigen values is illustrated in Figure 5. The scree plotindicates a single variable to be extracted as the eigen value of the first factor is 8.557and the second is 1.308. The difference between both factors is rather large and thechange in direction of the scree plot occurs after the first factor, thus indicating theappropriateness of extracting the first factor only.Figure 4.Mean and 5 per cent lowerconfidence limit ofmodel criteria

ECAM15,132Figure 5.Scree plot of eigen valuesfor criteria ofexcellence modelConfirmatory factor analysis is used to reassure the single factor structure asexpressed by the CFA Model in Figure 6, where the criteria of the ConstructionExcellence Model are related to a latent variable that is assumed to be “excellence”. Theterm “latent” refers to a variable that is not directly measured or is unobserved(Loehlin, 1998, p. 1). The CFA Model is expressed in the form of a path diagram withthe rectangles resembling items of the CFA Model and the latent variable portrayed asan ellipse. The arrows point from the latent variable to the items of the CFA Modelindicating that it is expressed in terms of the items (Cramer, 2003). A similarmanifestation of the EFQM Excellence Model as a CFA Model has been described inthe literature (Eskildsen et al., 2001, p. 788).The computations of the confirmatory factor analysis were conducted on theLISREL software. The observed variables (criteria) were obtained by the summatedscores of sub-criteria effectiveness. A program written in the SIMPLIS language wasdeveloped and the covariance matrix was prepared as input to the program using thePRELIS program that accompanies the LISREL package. The appropriateness of theCFA Model is determined by goodness-of-fit indices that indicate how well the data fitthe model. A large number of indices exist in the literature and there is little agreementas to which one is appropriate (Cramer, 2003). Of these indices, the root mean squarederror of approximation (RMSEA) is gaining popularity (Loehlin, 1998, p. 76) and isconsidered the only index to satisfy the ideal properties described by Gerbing and

A constructionexcellence model33Figure 6.CFA model of excellencecriteriaAnderson (1992) of indicating the degree of fit along a continuum with bounded values;being independent of sample size; and having known distributional characteristics.The value of RMSEA in the LISREL output for the CFA Model is 0.099, which indicatesa good fit since it is below the acceptable threshold of 0.1 (Cramer, 2003, p. 34; Loehlin,1998, p. 77). Therefore, the empirical data confirm the CFA model in Figure 6, whichillustrates the expression of the latent variable “excellence” in terms of the modelcriteria.Evaluation of model sub-criteriaThe sub-criteria of the Construction Excellence Model as developed from previousresearch are depicted in the sub-criteria of the questionnaire. These sub-criteria were

ECAM15,134assessed in the Questions 2 and 3 of the questionnaire as to their importance indefining their underlying criterion, and their actual effectiveness in the companies ofthe respondents, as can be seen in the Appendix (Figure A1). The sub-criteria of themodel are confirmed in a manner very much similar to that of the model criteria,through assessing their importance and through a factor analysis to ensure the singleconstruct structure.Importance of sub-criteria in defining underlying criterionThe results of the importance of each sub-criterion in defining its criterion in the formof the means, standard deviations and 5 per cent lower confidence limits werecomputed on an Excel spreadsheet, using its built-in functions in a similar manner tothe model criteria. The results show that the mean and lower confidence limits ofsub-criteria were all above average and inclined towards the scale maximum of fivepoints, hence confirming the importance of each sub-criterion in defining its underlyingcriterion. Furthermore, the negative skewness of these items indicate that even if a fewof these distributions are not normal, they are nevertheless peaked and negativelyconcentrated towards the upper end of the scale, as explained with the importance ofthe criteria previously. Thus, the actual confidence limits are expected to be higherthan those reported.Factor analysis of sub-criteria underlying a criterionThe results of the effectiveness assessment of each sub-criterion were used in factoranalysis to determine the single factor structure, which was previously conducted inthe discussion of the content validity and was demonstrated in Table II and Figure 3for the Leadership criterion. The factor analysis in all other criteria showed only onefactor is to be extracted and thus confirms the single factor structure assumed in thecriteria of the model.Calculation of model criterion weightsThe scoring system of the EFQM Excellence Model has been shown to vary acrossindustries and has been criticised as not corresponding to the way companies areworking (Eskildsen et al., 2001). Different methods for assessing the actual weights ofthe model’s criteria have been reported in literature. For example, Eskildsen et al. (2001)used factor regression coefficients. Cheng and Li (2001) used the analytic hierarchyprocess (AHP) to determine the weights of performance measures of a businessprocess. It is difficult to assess which method is more accurate. Nevertheless, any of theempirical approaches mentioned would be preferred to “an arbitrary weight structure,which has never been empirically tested” (Eskildsen et al., 2001, p. 783), as with thecase of EFQM. Furthermore, a method that has been previously applied to excellencemodels would be preferred. The factor regression coefficients method was chosen inthis research because it assesses the actual impact of each criterion on theorganisation’s performance, whereas with AHP requires expert judgement data thatcan be obtained from the importance rating of each criterion, but might affect theaccuracy of outcomes in the paired comparisons process, as well as reflecting theimportance perception rather than actual impact.

To compute the regression coefficients of the criteria relevant to one another, theyneed to be loaded on a single factor. Hence, the CFA Model of Figure 6, was used toreflect such a single factor structure. In this model, each of the criteria is loaded on alatent variable termed “excellence”. The regression coefficients are computed for eachcriterion in predicting the latent variable “excellence”. The computations were carriedout on SPSS, and the results are illustrated in Table III. Moreover, the coefficients intheir crude format might not be suitable for excellence scoring calculations and need tobe presented in a more workable format. Therefore, weights for each criterion werecalculated on an Excel spreadsheet such that the total weight of all criteria is 1,000.This was achieved by dividing the coefficient of each criterion by the total of allcoefficients and multiplying the result by 1,000.A constructionexcellence model35Benefits and limitationsThe main benefit of the Construction Excellence Model is that it provides an excellencemodel developed for and evaluated by construction companies. It has a morecomprehensive and wider coverage of performance criteria than other excellencemodels. Its underlying logic is also easier to understand and more user friendly. Thelimitation of the suggested model is that, as in all other models, it is only as good aswhoever uses it. Thus, the model is merely a tool that clarifies what needs to bemeasured/improved, but in no way can guarantee success of the organisation. Forexample, if management pursues an inappropriate strategy or the economy takes adownturn, the organisation’s performance would most certainly be affected, no matterhow efficient the business models it uses. However, the use of the suggested model isexpected to raise the chances of success and increase organisational performance inrelation to competitors. Furthermore, the use of the Construction Excellence Model, byits nature, is targeted towards project-based industries, specifically construction, andScore coefficientsCriterion weightsLeadership 0.0967 71Customer focus 0.0928 68Other stakeholder focus 0.0738 54Information and analysis 0.0877 65Strategic management 0.0939 69Intellectual capital management 0.0914 67People management 0.0903 67Partnership and supplier management 0.0952 70Resource management 0.0779 57Risk management 0.0890 66Process management 0.0712 53Work culture 0.0610 45Project performance 0.0833 61Organisational performance 0.0833 61Internal stakeholder performance 0.0778 57External stakeholder performance 0.0931 69Total 1.3584 1,000Table III.Component scorecoefficients andcriterion weights

ECAM15,136might not be applicable in other industries. This might limit the ability oforganisations to benchmark across industries.Conclusion and further workThe evaluation and analysis of the criteria and sub-criteria of a ConstructionExcellence Model was presented in this paper. The evaluation and analysis was basedon similar evaluations and analyses of TQM frameworks and excellence models inquality management literature. The analysis addressed the issue of data preparation,where missing data were analysed and appropriate data replacement methods used tocomplete the data set. The questionnaire as a measuring instrument was evaluated interms of reliability using Cronbach’s alpha, which resulted in the omission of the“Society and environmental impact of projects” sub-criterion from the “Projectperformance” criterion. Reliability of the measuring instrument is a necessary but notsufficient condition for validity. The validity of the measuring instrument is furtherevaluated in terms of content, construct and criterion-related validity using techniquesas factor analysis, multiple correlation. The criteria and sub-criteria were confirmedthrough their importance ratings and actual effectiveness ratings employingconfidence intervals and exploratory and confirmatory factor analysis. Theevaluation and analysis of the model’s criteria and sub-criteria strengthens themodel’s validity. The weights of the criteria were calculated using empirical data andthe method of factor regression coefficients.A number of future researches can be conducted in continuation of this researchpaper. The scope of the model could be expanded to suit the needs of other types ofconstruction organisations, such as consultants and owner organisations.Modifications could include, for example, modifying the relative criterion weights inthe Construction Excellence Model, for different types of organisations. Theconsideration of organisation size is another area of possible research, wherespin-off models could be developed for small and medium enterprises. In addition, thecriterion weights of national Excellence Models such as EFQM and Baldrige requireempirical justification. Finally, within the Construction Excellence Model, alternativemethods for computing criterion weights can be applied such as the analytic hierarchyprocess.ReferencesAhire, S.L., Golhar, D.Y. and Waller, M.A. (1996), “Development and validation of TQMimplementation constructs”, Decision Sciences, Vol. 27 No. 1, pp. 23-56.Ahmad, I.U. and Sein, M.K. (1997), “Implementing TQM principles in construction projects:difficulties and remedies”, International Conference on Leadership and Total QualityManagement in Construction and Building, CI Premier Conference, Singapore, pp. 47-54.Anderson, J.C., Rungtusanatham, M., Schroeder, R.G. and Devaraj, S. (1995), “A path analyticmodel of a theory of quality management underlying the Deming management model:preliminary empirical findings”, Decision Sciences, Vol. 26 No. 5, pp. 637-58.Bassioni, H.A., Price, A.D.F. and Hassan, T.M. (2004a), “Theoretical formulation of a frameworkfor measuring business performance in construction”, in Ruddock, L., Sexton, M.,

Amaratunga, D., Aouad, G. and Kagioglou, M. (Eds), 4th International PostgraduateResearch Conference, 1-2 April, University of Salford, Salford, pp. 419-30.Bassioni, H.A., Price, A.D.F. and Hassan, T.M. (2004b), “Performance measurement inconstruction”, Journal of Management in Engineering, Vol. 20 No. 2, pp. 42-50.Bassioni, H.A., Price, A.D.F. and Hassan, T.M. (2005), “Building a conceptual framework formeasuring business performance in construction: an empirical evaluation”, ConstructionManagement and Economics, Vol. 23 No. 5, pp. 495-507.Beatham, S.M., Anumba, C.J., Thorpe, A. and Murray, M.A.P. (2002), “Utilising the EFQMexcellence model to drive business improvement”, in Gancalves, A., Roy, M.A.P. andSteiger-Garcao, A. (Eds), Advances in Concurrent Engineering, 9th ISPE InternationalConference on Concurrent Engineering, Cranfield University, UK, A.A. BalkemaPublishers, Rotterdam, pp. 413-22.Black, S. and Porter, L.J. (1996), “Identification of the critical factors of TQM”, Decision Sciences,Vol. 27 No. 1, pp. 1-21.Brewerton, P. and Millward, L. (2001), Organisational Research Methods, Sage Publications,London.Byrne, B.M. (1998), Structural Equation Modeling with LISREL, PRELIS, and SIMPLIS: BasicConcepts, Applications and Programming, Lawrence Erlbaum Associates, Mahwah, NJ.Cheng, E.W.L. and Li, H. (2001), “Analytic hierarchy process: an approach to determine measuresfor business performance”, Measuring Business Excellence, Vol. 5 No. 3, pp. 30-7.Cooper, D.R. and Emory, C.W. (1995), Business Research Methods, Irwin, Chicago, IL.Cramer, D. (2003), Advanced Quantitative Data Analysis, Open University Press, McGraw-HillEducation, Maidenhead.Curran, P.J., West, S.G. and Finch, J.F. (1996), “The robustness of test statistics to non-normalityand specification error in confirmatory factor analysis”, Psychological Methods, Vol. 1,pp. 16-29.Duhachek, A. and Iacobucci, D. (2004), “Alpha’s Standard Error (ASE): an accurate and preciseconfidence interval estimate”, Journal of Applied Psychology, Vol. 89 No. 5, pp. 792-808.EFQM (2007), The EFQM Excellence Model, The European Foundation for Quality Management,available at: www.efqm.org (accessed July 2007).Eskildsen, J.K., Kristensen, K. and Juhl, H.J. (2001), “The criterion weights of the EFQMexcellence model”, International Journal of Quality & Reliability Management, Vol. 18 No. 8,pp. 783-95.Financial Analysis Made Easy (FAME) (2003), Company Database, Bureau van Dijk ElectronicPublishing, available at: http://fame.bvdep.com/ (accessed December 2003).Flynn, B.B., Schroeder, R.G. and Sakakibara, S. (1994), “A framework for quality managementresearch and an associated measurement instrument”, Journal of Operations Management,Vol. 11 No. 4, pp. 339-66.Gerbing, D.W. and Anderson, J.C. (1992), “Monte Carlo evaluations of goodness of fit indices forstructural equation models”, Sociological Methods & Research, Vol. 21, pp. 132-60.Gill, J. and Johnson, P. (2002), Research Methods of Managers, Sage Publications, London.Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W. (1998), Multivariate Data Analysis,Prentice-Hall, Englewood Cliffs, NJ.Hussey, J. and Hussey, R. (1997), Business Research: A Practical Guide to Undergraduate andPostgraduate Students, Macmillan Press, London.A constructionexcellence model37

ECAM15,138Kagioglou, M., Cooper, R. and Aouad, G. (2001), “Performance management in construction: aconceptual framework”, Construction Management and Economics, Vol. 19, pp. 85-95.Kaplan, R.S. and Norton, D.P. (1992), “The balanced scorecard – measures that driveperformance”, Harvard Business Review, January-February, pp. 71-9.Kaplan, R.S. and Norton, D.P. (1993), “Putting the balanced scorecard to work”, Harvard BusinessReview, September-October, pp. 134-47.Kaplan, R.S. and Norton, D.P. (1996), “Using the balanced scorecard as a strategic managementsystem”, Harvard Business Review, January-February, pp. 75-85.Loehlin, J.C. (1998), Latent Variable Models: An Introduction to Factor, Path, and StructuralAnalysis, Lawrence Erlbaum Associates, Mahwah, NJ.Makin, P.J., Cooper, C.L. and Cox, C.J. (1996), Organizations and the Psychological Contract:Managing People at Work, The British Psychological Society, Leicester.New Civil Engineer (2003), “NCE Contractors File”, New Civil Engineer, July.National Institute for Standards and Technology (NIST) (2007), Baldrige National QualityProgram, National Institute for Standards and Technology, available at: www.quality.nist.gov/ (accessed July 2007).Robinson, H.S., Anumba, C.J., Carrillo, P.M. and Al-Ghassani, A.M. (2005), “Businessperformance measurement practices in construction engineering organizations”,Measuring Business Excellence, Vol. 9 No. 1, pp. 13-22.Saraph, J.V., Benson, P.G. and Schroeder, R.G. (1989), “An instrument for measuring the criticalfactors of quality management”, Decision Sciences, Vol. 20 No. 4, pp. 810-29.Sekaran, U. (2003), Research Methods for Business: A Skill-building Approach, Wiley, New York,NY.Stevens, J.P. (2002), Applied Multivariate Statistics for the Social Sciences, Lawrence ErlbaumAssociates, Mahwah, NJ.Stockdale, D.J. (1997), “Can total quality management ‘add value’ in construction?”, in Harlow, P.(Ed.), Construction Papers, The Chartered Institute of Building, Ascot.Tabachnick, B.G. and Fidell, L.S. (2001), Using Multivariate Statistics, Allyn & Bacon, PearsonEducation, Needham Heights, MA.Tan, K.C. (2002), “A comparative study of 16 national quality awards”, The TQM Magazine,Vol. 14 No. 3, pp. 165-71.Watson, P. and Seng, L.T. (2001), “Implementing the European Foundation for QualityManagement Model in construction”, Construction Information Quarterly, Constructionpaper 130.Wilson, D.D. and Collier, D.A. (2000), “An empirical investigation of the Malcolm BaldrigeNational Quality Award causal model”, Decision Sciences, Vol. 31 No. 2, pp. 361-90.Further readingKelloway, E.K. (1998), Using LISREL for Structural Equation Modeling: A Researchers Guide,Sage Publications, Newbury Park, CA.

Appendix.A constructionexcellence model39Figure A1.Questionnaire

ECAM15,140Figure A1.

A constructionexcellence model41Figure A1.Corresponding authorH.A. Bassioni can be contacted at: hbassioni@yahoo.co.ukTo purchase reprints of this article please e-mail: reprints@emeraldinsight.comOr visit our web site for further details: www.emeraldinsight.com/reprints