5:23 0 0% A Call 90% uncertainty. comprising project success criteria. Subsequently, Stefanovic (2007) argued that teamwork effect Similarly, Bryde (2003a) linked TOM and PM practices iveness should be included with these dimensions. Teamwork proposing a model based on the EFQM model called the effectiveness is considered as a component of project success in "Project Management Performance Assessment (PMPA)'. many other studies (Bryde, 2008; El-Saboni et al., 2009; Lim and Instead of the nine criteria used in the EFQM model the Mohamed, 1999; Muller and Jugdev, 2012; Muller and Turner, PMPA model consists of five enablers of high PM perfor- 2007; Pinto and Pinto, 1991; Wateridge, 1998 and Westerveld, mance; PM leadership, PM staff, PM policy and strategy, PM 2003). The model presented by Shenhar et al. (2001) was selected partnerships and resources and project life cycle management for investigation in our study and was adapted to include teamwork process. The final area in the PMPA is PM Key Performance effectiveness. Indicators (KPIs), which are the practices by which actual achievement is measured. The PMPA model is presented in 2.2. PM Performance Fig. 1. Bryde (2003a) explains the integral parts of the PMPA model as follows: Traditional PM systems which exclusively pursue the success criteria of cost, time, quality and meeting technical 1. PM Leadership: Focuses on requirements have become considered ineffective (Bourne et development and promulgateon of awareness of the role of al., 2000; Walton and Dawson, 2001). A common approach is projects as a vehicle for managing all types of change to focus on multiple stakeholders' expectations (Bryde, 2003b; . ensuring that PM system supports the development of open, Maylor, 2001; Tukel and Rom, 2001). This has led to a new set wo-way partnerships with customers and suppliers and a of difficulties in developing models for measuring perfor- shared, common project language culture. mance because stakeholders' needs are often difficult to 2. PM Staff: Emphasises manage and measure (Boehm and Ross, 1989; Maylor, 2001) . the planning and management relating to PM staff to and there is sometimes resistance to going beyond the increase its PM capability by maximising the potential of traditional criteria due to commercial pressures (Chan et al., project-related human resources 2003). These difficulties have resulted in limited literature on . the extent to which the management of PM staff more holistic performance assessment frameworks for project incorporates methods for rewarding performance relating environments. to PM. It is evident in the literature that TOM and PM are two key 3. PM Policy and Strategy: Focuses on how the development management approaches implemented in organisations for achiev- of PM, across an organisation, is introduced in a planned and ing continuous improvement and organis nal success (Bryde, systematic fashion ensuring the linkage between strategic, 1997). A positive correlation has been found between TOM organisation level and the tactical, project level. practices and organisational performance (Barad and Raz, 2000; 4. PM Partnerships and Resources: Emphasises: Broetzmann et al., 1995; Choi and Eboch, 1998,). Similarly PM is . The role and importance of win-win partnerships found to be an effective tool for achieving the strategic objectives between all stakeholders of organisations (Kerzner, 2003), managing organisational change . Effectiveness of such partnerships on project manage- Bryde, 2003b; Maylor, 2001) and systematic planning, execution ment strategy and control of activities in a systematic mar er (Meredith and 5. Project Lifecycle Management Processes: Incorporates pro- Mantel, 2003). cesses which are required to manage the whole project life cycle The literature indicates that a two-way linkage exists 6. PM Key Performance Indicators (KPIs): Focuses on between PM and TOM (Broetzmann et al., 1995; Choi and . KPIs to indicate results achieved in relation to meeting the Eboch, 1998). PM is recognised as an effective methodology requirements of project stakeholders for implementing TOM practices in organisations. Similarly, . The methods used within the PM system to improve TOM plays a role in providing an environment which facilitates performance against the KPIs. 3/16 F.A. Mir. A.H. Pinnington / International Journal of Project Management 32 (2014) 202-217 205 PM Staff Project Project Life (PM) Management PM Key Performan PM Policy & Strategy Processes Indicators (KPIs) PM Partnerships & Resources ENABLERS RESULTS Fig. 1. The PMPA model (Bryde, 2003a, p. 233). After detailed study of Westerveld's (2003) Project Excellence The following hypotheses will be tested to test the relevance Model and Bryde's (2003a) PMPA model, the PMPA model was of this proposition; selected for this study as the preferred model for the performance assessment of PM. EFQM is a tried and tested model and PMPA Hypothesis 2. H2 has a closer resemblance to this model while being comprehen sive enough to cover all of the aspects identified in the Project There is a statistically significant positive relationship Excellence Model situated in the PM context. Furthermore, between PM Leadership and Project Success. though Westerveld (2003) advocated the use of his model based on critical success factors and project success criteria, he did not Hypothesis 3. H present any statistical background to support his claim. In contrast, the PMPA model has been validated by Qureshi et al. There is a statistically significant positive relationship (2009) and Din et al. (2011). between PM Staff and Project Success. Hypothesis 4. HA 2.3. PM Performance and its relationship with Project Success There is a statistically significant relationship positive The PM literature argues that there is a positive relationship between PM Policy and Strategy and Project Success. between PM Performance and Project Success (Bryde, 2008; Hypothesis 5. Hs Munns and Bjeirmi, 1996) and Munns and Bjeirmi (1996) claim that Project Success is dependent on appreciation of the There is a statistically significant positive relationship importance of PM. They further emphasise that this role must between PM Partnerships and Resources and Project Success. be considered in terms of the wider organisational strategy and long-term expectations. From the above discussion it has been Hypothesis 6. He argued that Project Success and PM Performance are distinct yet inter-related concepts and a positive relationship between There is a statistically significant positive relationship between them is sought. So, it is proposed that: PM Project Lifecycle Management Processes and Project Success. Hypothesis 7. H7 Proposition 1. There is a positive influence of Project Management Performance on Project Success (Bryde, 2008; De Wit, 1988; Jugdev and Muller, 2005; Morris, 1998; Munns There is a statistically significant positive relationship and Bjeirmi, 1996). between management of Key Performance Indicators (KPIs) and Project Success. Hypothesis 1. H, One school of thought argues that researchers should acknowledge the multi-dimensional nature of Project Success There is a positive statistical relationship between PM (Atkinson, 1999; Lim and Mohamed, 1999; Lipovetsky et al., Performance and Project Success. 1997; Shenhar et al., 2001; Stefanovic, 2007), and based on Adopting the PMPA framework, the second proposition of Propositions 1 and 2, this study considers Proposition 3 which is: O5:23 0 09 . Call 90% 0.4) for all Project Success considered acceptable in social science research (Cronbach, variables. 1951; Nunnally, 1978). All the alphas for variables within Project Success were above the acceptable value of 0.7. 4.2.2. Linear regression of independent variables with Project The final results were: PM Performance construct (92.7%) Success construct and Project Success construct (93.2%). Cronbach alpha of the To further explore the collected data and validate earlier overall survey tool after excluding Q7 & Q8 was 95.4%. These inferences about correlations, linear regression tests were results confirmed the appropriateness of further analysis of the conducted. Summarised results of linear regression are data without any further deletion of items. presented in Table 3 whereas the first column of this table gives the reference to corresponding Research Hypothesis. 4.2. Inferential Statistics; Correlations and Regressions Some key findings from Table 3 are as follows: The sample consisted of 154 responses. According to the a) PM Performance explained 44.9% of the variance in central limit theorem, for such sample sizes, the sampling Project Success, with a very significant relationship distribution will take the shape of a normal distribution (Field, explained by F values and Beta values (F = 125.47, B = 2009). This is reasonable justification for using parametric tests 0.672, p Notes: Correlation is significant at the 0.01 level (2-tailed) e 0.01 level (2-tailed) for all variable associations in the table given above. Key findings are as follows: 4.2.5. Multiple regressions A stepwise approach using both forward and backward ) Table 5 lists 7 model summaries (first 6 models representing methods of regression was used. A best fit model was generated each variable of Project Performance and 7th model was by backward stepwise method, explaining most variance in the made for Project Performance construct itself) for each dependent variable (Project Success). The results are given in variable of Project Success construct (dependent variables). Tables 6 and 7. These tables indicate that Model 3 was the best fit; ) Table 5 also highlights top 10 (solid shaded boxes) and PM Leadership, PM Staff, PM Partnerships & Resources and PM bottom 10 (shaded with diagonal lines) model fit values KPIs collectively define 45% variance in Project Success. Adjusted R Square) to indicate which Independent Variables Adding or removing any further independent variable decreases explain most in the corresponding Dependent Variables. the model fit. ) The table with its legend mentioned above indicates that Project Performance construct itself is amongst the top 10 4.3. Construct validity models hence explaining the most for each variable of Project success, except Project Efficiency. It explains the Principal Component factor Analysis (PCA) with varimax most in Impact on Project Team (42.3%) and the least for rotation was conducted to validate the underlying structure of Project Efficiency (17.2%). the PM Performance and Project Success constructs. ) Impact on Project Team is the single-most-variance-explained Project Success variable by the majority (4 out of 6) of Project 4.3.1. PM performance construct validity Performance variables. PM KPIs explains the most variance PCA was employed for factor extraction on the 20 items of (32.5%) after Project Performance explained above. The same PM Performance construct. Kaiser (1960 in Field, 2009) result was earlier found during correlation of the Project recommends retaining all factors with eigenvalues greater Success variables with Project Performance variables indicat- than 1. The results revealed the presence of four components ing the fact the Impact on Project Team has the highest with eigenvalues exceeding 1, explaining 42.9%, 7.3%, 5.8% correlation with each variable of Project Performance. and 5.6% of the variance respectively. After Varimax rotation ) Impact on Project Team is followed by Preparing for Future and the four-component solution explained a total of 61.5% of the Impact on Customer each having 2 variables of Project variance, with component 1 contributing 27.5%, component 2 Performance among top 10 model fit values. Similar results contributing 13.85%, component 3 contributing 11.09% and were earlier found during correlation of Project Success component 4 contributing 9.07%. variables with Project Performance variables. The results show comparatively higher correlation of Preparing for Future 4.3.2. Project Success construct validity and Impact on Customer with each Project Performance PCA revealed the presence of four components with variables compared to Busines and Project Efficiency. eigenvalues exceeding 1, explaining 44.6%, 8.96%, 6.76% and Project Efficiency is the least-variance -explained dependent 5.12% of the variance respectively. To aid the interpretation of variable with 5 out of 6 Project Performance variables among these four components, Varimax rotation was performed. The the bottom 10 model fit rates among all possible models. four-component solution explained a total of 65.4% of the Maximum variance explained by any model for this variable variance, with component 1 contributing 22.78%, component 2 was by PM Performance itself with 17.2% model fit and it was contributing 18.55%, component 3 contributing 12.82% and followed by PM KPI with model fit value of 16.8%. A similar component 4 contributing 11.26% result was found in the correlation test of Project Success with Project Performance variables indicating that Project Efficien- 5. Discussion cy had lowest correlation with most of the Project Performance variables. 5.1. The PM Performance-Project Success relationship g) Business Success was also found to be least explained by almost all variables of Project Performance with most The statistically positive relationship found between PM variance being explained by the model with PM Perfor- Performance and Project Success is consistent with previous mance itself (25%). research (Din et al., 2011; Stefanovic, 2007; Stefanovic and 8/16 210 F.A. Mir, A.H. Pinnington / International Journal of Project Management 32 (2014) 202-217 1.715 1.759 1.976 1.899 2.604 2.556 results for correlation test and linear regression tests. endent variables Dependent Pearson's Linear variable Correlations regression idardised Tolerance VIF 301 results Sig. HI PM Performance Project Success .672 125.47 0.000 fficients 1 H7 PM KPIs Project Success .578 76.178 0.000 O5:23 0 09 . Call 90% short-term benefits (for example, meeting cost, time, quality Table 3 HI HS objectives of current projects), but also the long-term 9/16 F.A. Mir. A.H. Pinnington / Intern ional Journal of Project Management 32 (2014) 202-217 211 Table 5 Linear regression results for independent variables with individual variables of Project Success. Independent variable R Square Adjusted R square Std. error of the Dependent estimate variable a PM Leadership .438(a) 0.192 0.187 4.014 C PM Staff 492(c) 0.242 0.237 3.889 d PM Policy and strategy 406(d 0.165 1 159 4.081 e PM Partnerships and resources 443(e) 0.197 0.191 4.003 Impact on customer f PM Lifecycle management processes 470(f) 0.221 0.216 3.943 g PM KPIs 444(g) 0.197 0.192 4.002 h PM Performance .560(h) 0.314 0.31 3.699 a PM Leadership 554(a) 0.307 0.302 2.127 C PM Staff 546(C) 0.298 0.294 2.141 d PM Policy and strategy 435(d) 0.189 0.184 2 301 Impact on project team e PM Partnerships and resources 149(e 0.201 0.196 2.283 f PM Lifecycle management processes .537(f) .288 0.284 2.156 & PM KPIs .574(8) 0.33 0.325 2.092 h PM Performance 653(h) 0.426 0.423 1.935 a PM Leadership .359(a) 0.129 2.68 c PM Staff 451(c) 0.203 0.198 2.564 d PM Policy and strategy 405(d 0.164 8 159 2.625 e PM Partnerships and resources 376(e) 0.141 2.661 Business success f PM Lifecycle management processes 401(1) 0.161 2.631 g PM KPIs 438(g) 0.192 0.186 2.582 h PM Performance 506(h) 0.256 0.251 2.477 a PM Leadership 369(a 0.136 2.685 c PM Staff 365(c) 0.133 2.69 d PM Policy and strategy 229(d 2.813 Project efficiency e PM Partnerships and resources 254(e) 2.794 f PM Lifecycle management processes .332(f) 2.726 8 PM KPIs 417(8) 0.168 2.626 h PM Performance 421(h) 0.177 0.172 2.621 a PM Leadership 459(a) 0.211 0.206 2.268 C PM Staff 431(C) 0.186 0.181 2.304 O5:23 0 09 . Call 90% The linear regression results showed that: shows the need for keeping all the project relevant staff informed about the Business Success that is being achieved a) Impact on Project Team is the single-most-variance-explained through individual projects. The result indicating Project Project Success variable by the majority (4 out of 6) of Project Efficiency and Business Success being least explained by all Performance variables. The same result was earlier found variables of Project Performance deserves additional inves- during correlation of Project Success variables with Project tigation as that does not conform to generally believed PM Performance variables indicating that the Impact on Project practices. Team has the highest correlation with each variable of () An organisation's future success (represented by Preparing Project Performance. The results show the importance of the for Future) is greatly impacted by lifecycle management Impact on Project Team variable for PM practitioners. Project processes and systems implemented in the organisation as performance can have a major impact on project teams. The shown by the results. This provides empirical evidence of perception of a successful project motivates the team and the long-term benefits that an organisation can achieve by increases team member engagement and commitment to the investing in the lifecycle management processes and project as well as to the team itself. It is also seen that PM KPIs systems within the organisation. explains the most variance (32.5%) among the individual PM Performance variables in Impact on Project Teams. This result 5.4. Validation of the PMPA framework and project success showing PM KPIs as being the most effective variable to construct explain variance in Impact on Project teams highlights an important finding for Project Managers and HR Managers, This study also explored the validity of the variables/factors i.e., having an effective PM KPI management framework that were proposed by the literature in the context of this study positively and highly influences the engagement of project (i.e. project-based organisations in the UAE). teams. An exploratory Principle Component Factor Analysis was b) PM KPIs seem to have the most wide-ranging impact across performed for PM Performance and Project Success scales the different variables of Project Success. It has the highest to determine the resultant factors. These modified PM Perfor- correlation with Impact on Project Teams, followed by Project mance and Project Success constructs were renamed as PM_ Efficiency, Preparing for Future and Business Success, in the Performance_Modified and Project Success Modified. From same rank order. Therefore, it is concluded that having a the results re-organisation of PM performance factors was formal management system for developing, managing and made, shown in Table 8. updating KPIs formally in an organ ation can directly impact Examining the PM Leadership Component loadings it can on Team Performance and Project Efficiency. be interpreted that some items relevant to PM culture did not ) The results also indicate that in addition to PM KPIs, Ps . ctually belong to this scale. It was noticeable that the PM Lifecycle Management Processes and PM Staff ar dership scale had the lowest reliability initially ( e. A limitation of this study is it is based on self-report f Project Management 17, 337-342. responses and such responses are often known to be affected Avots, I., 1969. Why does project management fail? California Management by participants' biases. It is possible that bias was introduced Review 12 (1), 77-82 (pre-1986) as participants, in their retrospective assessments, cannot or Barad, M., Raz, T., 2000. Contribution of quality management tools and do not, always accurately recall a past situation's attributes. practices to project management performance. International Journal of Quality & Reliability Management 1 Management 17 (4/5), 571-583. Future studies could be designed to have two (or more) Belassi, W., Tukel, O.I., 1996. A new framework for determining critical perspectives from each organisation. success/failure factors in projects. International Journal of Project f. The survey was dependent on only one group of respondents Management 14 (3). 141-151. for the independent an pendent variables (self-report esner, C., Hobbs, B., 2006. The perceived value and potential contribution of project management practices to project success. Project Management responses). The survey data collection therefore suffers from Journal 37 (3), 37-48. the common method variance problem (Podsakoff et al., Boehm, B.W., Ross, R., 1989. Theory-W software project management: 2003). To eliminate the occurrence of response bias, future principles and examples. IEEE Transactions on Software Engineering research could collect data from other relevant project (7), 902-916. stakeholders, particularly, project owners, executive direc- Bourne, M., Mills, J., Wilcox, M., Neely, A., Platts, K., 2000. Designing implementing and updating performance measurement systems. Interna- tors, and project steering groups. ional Journal of Operations & Production Management 20 (7), 754-771. amann, S.M., Kemp, J., Rossano, M., Marwaha, J., 1995. Customer 9. Conclusion sfaction -lip service or management tool? Managing Service Quality 5, This research study demonstrates that PM perform. A., Adams, J., 2000. Measuring the effect of project management on correlated to Project Success within UAE organisation. astruction outputs: a new approach. International Journal of Project Management 18, 327-335. paying greater attention to this relationship, organisations Can Bryde, D.J., 1997. Underpinning modern project management with TOM increase their rate of project success. principles. The TOM Magazine 9 (3) It was seen through linear regression analysis that PM Bryde, D.J., 2003a. Modelling project management performance. International Performance explains at least 44.9% variance in Project Success. Journal of Quality and Reliability Management 20 (2), 229-254. O