Measuring the Characteristics among Critical Success Factors of PPP Infrastructure Projects

Private-Public Partnership (PPP) is an increasingly popular choice for policymakers in implementing critical public projects through the examination of essential factors of success of PPP and establishing an empirical model of PPP in the construction project in Malaysia. The PPP implementation model was hypothesised to investigate the measurements and dimensions of technology, Organisational, and project characteristics as critical success factors of PPP implementation. A quantitative methodology was employed to validate the measurements, hypothesis testing and validate a structural model of PPP implementation. A total of 238 respondents was involved in the survey of the hypothesised PPP model. SPSS version 22 as well as Analysis of Moment Variance (AMOS) software were employed to analyse the data gathered. Path analysis and mediated regression analysis of the structural equation model succeeded in determining the mediating effect of stakeholder and procurement on the relationship among critical success factors and PPP implementation. The overall results show a significant positive interaction of Organisational, technical, and project characteristics as essential factors of success on stakeholder and procurement as a mediator on the achievement of PPP implementation. This paper highlights not only the vital success factors for PPP but also offers a fundamental contribution model achieved through the empirical model of critical success factors and PPP implementation in the construction project in Malaysia. This study succeeds in establishing and validating a structural model of the PPP implementation model. The model contributes to the body of knowledge of PPP and benefits to practitioners as primary guidance on construction and business developments


INTRODUCTION
The PPP is a common choice for policymakers to undertake significant projects in public works.It is a technique and framework for managing project management of infrastructures (Solla et al., 2019); however, faced with the dearth of government savings and where general inefficiency is to be combated.An example of that is the local private partnership in the Russian Federation and alteration of separate legislative acts (Berezin et al., 2018).In the end, PPP has a better value for investment from public sector services (Casady et al. 2019).The recent achievements with the PPP model in Hong Kong, Singapore, Australia, UK, and other nations have made PPP increasingly popular in Malaysia.This research focuses on the findings of the recent PPP project survey in Malaysia on variables that are considered crucial to the PPP project's success.Recent work primarily has several aims.First, it analyses the importance of the results of the variables interpreted by the respondents.Second, the study looks at discrepancies in the awareness of the significance of output factors between the private and public sectors (Ismail, 2013).The previous study additionally indicated that the PPP method is gaining acceptance worldwide.Thus, few studies attempt to examine the operational and managerial dimensions of PPP from an accounting perspective in the context of developed nations (Cheung et al., 2012).It is, therefore, opportune to discuss this area.Critical success factors are correlated with practical implications for a company that can support an organisation's functioning and longevity.
The concept of critical success factors (CSFs) was also developed (Joreskog, 1984).(Saraph et al., 1989, Saron et al., 2013), and it is seen CSFs as 'the critical areas of managerial planning and action that need to be implemented for success.It is cantered on (Pfeffer and Sutton, 2000, Rockart, 1979. Schwandt, 1999) the comprehension of corporate executives strengthened by the awareness of CSFs about their processes to reduce the price of job failure.The notion of CSF is cutting through various fields of human activities (Ansarinejad et al., 2011), where process creation is needed.
Besides (Morledge and Owen 1999a, Morledge and Owen 1999b), the notion of CSFs was developed to define obvious vulnerabilities linked to the functional implementation of Rockart's method.However, it involves bias, subjectivity, human inability to interpret complex information, changes in the environment, generalisations, and imprecise definitions and qualitative outcome measures.This cycle of CSFs has been utilised as a control measure in many sectors.As a consequence, attempts to apply this same principle to construction management have been made.Cheung (2009) used 18 variables to analyse PPP opinions on CSFs.The elements were decomposed into five classes or factors that underlie them: stable macroeconomic environment, the shared responsibility between the public and private sector, effective and transparent procurement procedure, the firm social and political context, and reasonable government control.PFI (Rashid, 2014).The latter issues are shifting responsibility for funding and maintaining a package of capital expenditure and utilities to the private sector including construction, management, repair, renovation, and replacing public sector (Alfen et al., 2009).The lease payment arrangement for PFI ventures would ensure a complete return on the concessionaire's capital investment costs consisting of repayment of expenditure and investment income.
The properties and services will be returned to the public sector upon the expiry of the concession agreement (Muhammad and Johar, 2018).In Malaysia, through the selling of the equity process, Cost and Project (Azadegan et al., 2013, Ismail, 2013, Ismail and Azzahra Haris, 2014, Ismail and Rashid, 2007, Ismail et al., 2009) The research design is deemed highly significant in both data collection and analysis stages with a plan for achieving the objectives of the investigation (Oppenheim, 2000).The research design ought to contain clear goals obtained from the research questions.It should also identify the sources from which the researcher intends to collect data.The following series of rational decisions are involved in creating the research design: identifying the aim of the research whether it is preparatory, hypothesis testing or descriptive; identifying the type of investigation; deciding the extent of intervention by the researcher; identifying the study setting; deciding on measurement and measures; deciding on data analysis; data collection methods, time horizon; sampling design; and identifying the analysis unit.This study utilises a quantitative approach using the questionnaire for an investigation of the relationship among dimensions of PPP implementation.However, that provides an understanding of the current construction factors affecting the stakeholder and procurement as mediating variables among the sample under study.
After the investigation has identified both the existence and characteristics of the elements, it determines the positive attributes that contribute to the successful implementation of PPP in the Malaysian construction industry.

QUANTITATIVE RESEARCH FRAMEWORK AND HYPOTHESIS
The survey and questionnaires have been widely used to measure the relationship between PPP factors, stakeholder, procurement, and the effect of PPP implementation.This survey data can be collected using various methods that are compiled in either a single fashion or in combinations.Considering all the theoretical underpinnings required for this research, the researcher has decided to use a survey method to effectively extrapolate information and analyse it against a given hypothesis based on PPP. Figure 1 and Table 3 present the research framework and hypothesis.Training confirms as a significant dimension of an Organisational factor.Positive Hy.2 Time, cost, and schedule confirm valid and significant measurements of project factor of PPP.

Positive
Hy.2a Time confirms as a significant dimension of the project factor.Positive Hy.2b Cost confirms as a significant dimension of project factor.Positive Hy.2c Schedule confirm as a significant dimension of the project factor.Positive Hy.3 Perceived usefulness and perceived ease of use confirm the as valid and significant measurement of a technological factor of PPP.

Positive
Hy.3a Perceived usefulness confirms as a significant dimension of technological factor Positive Hy.3b Perceived ease of use confirms as a significant dimension of technological factor Positive Hy.4a Organisational factor has a direct positive significant influence on stakeholder.Positive Hy.4b Organisational factor has a direct positive significant influence on procurement.Positive Hy.5a The project factor has a direct positive significant influence on stakeholders.Positive Hy.5b The project factor has a direct positive significant influence on procurement.Positive Hy.6a The technology factor has a direct positive significant influence on stakeholders.Positive Hy.6b The technology factor has a direct positive significant influence on procurement.Positive Hy.7 Stakeholder has a direct positive significant influence on PPP implementation.Positive Hy.8 Procurement has a direct positive significant influence on PPP implementation.Positive Hy.9 Organisational, project, and technological factors, mediated by stakeholder and procurement, have a significant interaction on PPP implementation in the valid structural PPP model.

Positive FINDING AND DISCUSSION
In this section, the profiles analytic approach was utilised to study the mass of quantitative data, which was recorded over a period.In an attempt to provide meaning, order, and structure to the information, the investigators tried giving a thorough discussion of the analysis process.The data analysis assimilated the emergent themes and the reality of the phenomena investigated.This investigation is geared towards establishing the idea that data capture's form is essential via text.Most information was turned into a document which then served as the primary model for the interpretation of the object (Schwandt, 1999).As highlighted in the preceding section, the information was organised via various approaches in the final analysis.This version of information serves as the integral component of a procedure that involves the simultaneous and sequential recording and study of data (Creswell, 2002).Following that, all documented sources have undergone expansion on the data analysis approaches.This study applies to SPSS, Amos™ 22, and theory approach by Charmaz (2006) andTosolini et al. (2014).
Instead of constructing a method, the goal was to provide a viable interpretation of the outcomes.The next sections describe the detailed steps of data analysis.The package connects the information to the specified model and generates results including parameter projections and overall model fit statistics.According to Byrne (2013) and Byrne (2016), the analysis's input is a covariance matrix of measured variables including survey item scores.However, occasionally the models of covariance and means or patterns of correlations are utilised.In practice, SEM programmes with the necessary information provided by the data analyst transform this information into means and covariance to be used.As shown in Figure 2, the simple methodology in running an SEM analysis by Ferdinand (2002).The results feature the model's overall indices, standard errors, test statistics, and parameter projections for each free variable in the model.The SEM possesses some good merits to configure PPP and models of critical success factors (Ferdinand, 2002).FIGURE 2. SEM Procedure (Byrne, 2001) The model includes relationships among the measured variables.These relationships are then expressed as restrictions on the entire set of possible relationships in Figure 3.The overall model fits indices and parameter estimates, standard errors, and test statistics for each free parameter in the model.SEM design of PPP and vital success factor models have many attractive merits (Ferdinand, 2002).

SUMMARY OF PRELIMINARY RESULTS
Following the data analysis of actual data collection from 285 respondents, the multiple-item measures underwent a series of reliability and validity checks.
The tests were all valid, internally consistent, unidimensional, and reliable.An oblique rotation (Direct Oblimin) procedure with PCA was utilised to generate the measures' dimensionality.The steps were all found to exhibit outstanding reliability with coefficient alpha within the range of 0.60 -0.90.This particular range is within the scope of an acceptable level of 0.60.All the measures were unidimensional and passed a factor analysis with exceptional loadings (Sekaran and Bougie, 2016).All the rules were unidimensional, reliable, and valid.Their performance was also excellent and following the outcomes from the pilot test and pretest.

CONFIRMATORY FACTOR ANALYSIS OF MEASUREMENTS
The measurement of the hypothesised in the structural model was confirmed via Structural Equation Modelling.SEM is a statistical method that is capable of assimilating the structural model and measurement model or CFA into simultaneous statistical investigations (Byrne, 2013).SEM is especially vital in hypothesis testing and inferential data analysis, in which the pattern of interrelationships between the constructs is assigned a priori and based on established theory (Byrne, 2002).It possessed the ability to model links among criterion parameters and multiple predictors as well as run statistical investigations on a priori theoretical assumptions against practical information via CFA (Sentosa and Mat, 2012).In this particular investigation, the measurements are compared with the hypothesis in the structural model using CFA.

CFA OF TOP MANAGEMENT SUPPORT (ORGANISATIONAL DIMENSIONS)
The confirmatory factor analysis (CFA) of the five items of (TSM) Top Management Support is presented in Table 4.Moreover, the result confirms that all five items are validated Figure 4.The estimation of factor loading for Management number one: "Top management provides sufficient resources for the implementation of public-private partnerships." is 0.690, which shows this item is validated.The significant level 0.000 proves that it is substantial.Management number two: "Top management provides stable funding for the implementation of public-private partnerships." is validated at the estimation of 0.881, and it is also significant.Management number three: "Top management regards the implementation of public, private partnerships as a high priority."It has an estimation of 0.819 and a significance level of 0.000, which confirms it is both validated and significant.
Management number four: "Top management provides constructive feedback on the appropriateness of the implementation of public, private partnership "is verified at the level of 0.952, and it is also significant.Moreover, for the last item, Management number five, "Top management encourages the staff to participate in the implementation of public-private partnerships, "the factor loading is 0.866, which confirms its validity, and due to its significance level 0.0000, it is also found significant.5 represents the goodness of the fit indexes of the Top Management Support.The ratio is less than 2 and having a value of 0.655 shows that the model is fit.The P-value is with a ratio of more than 0.05 and at 0.623 value, and GFI is more than 0.9.The value was 0.996, TLI with a ratio of more than 0.9 and a value of 1.004.Finally, RMSEA's critical rate of less than 0.08, and the importance of 0.000 altogether confirms that the model of top management support is fit and significant.5. Item one, our Company provides internal training on public-private partnership activities, and it has a factor loading of 0.823 and a significant level of 0.000; hence, it is validated and significant.Item two: "Our Company provides clarity of staff role and objectives before training regarding public, private partnership activities."has the factor loading 0.823 and significance level of 0.000, which makes it both validated and significant.Moreover, the third item: "Our Company provides adequate course material during the training of public-private partnership activities."has an estimation of 0.770 and a significance level of 0.000.This item is also both significant and validated.Item four: "Overall, I am satisfied with training sessions regarding publicprivate partnership activities" has factor loading of 0.791 and 0.000 level of significance, which again makes it both significant and validate.The P-value with a ratio greater than 0.05 and a value of 0.108, GFI, more significant than 0.9 at the value was 0.991, TLI with a ratio greater than 0.9 and a value of 0.985.Finally, RMSEA's critical ratio of less than 0.08 and the importance of 0.072 altogether confirm that the model of training is fit and significant.Table 8 shows the Confirmatory Factor Analysis (CFA) of the four items of Support.The result confirms that all four items are validated in Figure 6.The factor loading for item one: "Our Company provides guides from the technical centre for publicprivate partnership activities "is 0.799 at the significance level 0.000, which shows the first item is validated and is significant.The estimate for the second item: "Our Company assigns a specific person (or group) for assistance with difficulties encountered public-private partnership activities." is 0.709, and P is 0.000, which means this item is also validated and significant.For the third item: "Our company provides specialised and customised instructions concerning popular and useful publicprivate partnership activities." the factor loading is (0.706), and P is 0.000, which makes this item validated and significant as well.Item number four: "Overall, I am satisfied with the performance of the technical centre regarding public-private partnership activities."has an estimate of (0.552) and the P shows 0.000, which confirms this item to be validated and significant.

TABLE 8. CFA Support
Table 9 shows the goodness of the Fit indexes of Support.The ratio is less than 2, and the value being 2.066 confirms that the model is fit.Also, the P-Value ratio more than 0.05, along with the Value 0.127, GFI more than 0.9, and its Value 0.991, TLI being more than 0.9 with the Value 0.976, and finally RMSEA being less than 0.08 with the Value 0.067 confirm that the model is fit and significant.Table 10 shows the Confirmatory Factor Analysis (CFA) of the four items of Perceived Usefulness Figure 7.The result confirms that the four items out of five are validated and significant.The factor loading for item one: "I feel using technology improves the quality of work of public-private partnership implementation" is 0.800, with P being 0.000, which makes the first item validated and significant.Item two: "I feel using technology gives me greater control of public-private partnership implementation."at the estimate of 0.703, and a significance level of 0.000 was validated and significant as well.Item three: "I feel technology enables me to do my job more quickly."with a factor loading of 0.683, and the P 0.000 was confirmed to be validated and significant.Furthermore, item four: "I feel using technology enhances my effectiveness of public-private partnerships implementation."with the factor loading of 0.676 and the P 0.000 was found validated and significant as well.One item was removed.Table 11 shows the goodness of the fit indexes of Perceived Usefulness.The ratio is less than two and the value 0.089 demonstrates that the model is fit.The P-value with ratio more than 0.05 and value of 0.915, GFI with ratio more than 0.9 and value 1.000, TLI with ratio more than 0.9 and the Value of 1.019 and RMSEA with the ratio of less than 0.08 and value of 0.000 confirm that model is fit with significant.Table 12 shows the Confirmatory Factor Analysis (CFA) of the five items of schedule.The result confirms that all five items are validated in Figure 8.The factor loading for item one: "the PPP project schedules were strictly adhered to" stands at 0.834 with the P 0.000, which confirms the validation and significance of the first item.For item two: "overtime is better at meeting PPP project cost," the factor loading 0.811 and the P0.000 show that it is both validated and significant as well.

Index
The factor loading for the third item: "PPP project rarely deviates from settle plans" is 0.865, with the P being 0.000, which confirms the significance and validation of the item.For the fourth item: the PPP project relaxes deadlines to meet schedules fully, the factor loading is 0.802, and the P is 0.000 proving it to be validated and significant.Finally, the factor loading is 0.778 for the last item "Overall project schedule performance was met based on baseline goals, targets, or expectations."The significance level shows that this item is both validated and significant as well.Table 13 shows the Confirmatory Factor Analysis (CFA) of the six items of cost.The result confirms that all six items are validated in Figure 9.The factor loading for item one: "Our Company insists on meeting PPP project cost" is 0.581, with the P being 0.000, which makes the item significant and validated.For item two: "Our company over time is getting better at meeting cost" the factor loading is 0.584 and the significance is 0.000, which confirms the validation and importance of it.The loading factor of item three: "Our Company takes corrective action to control PPP project costs," is 0.513, and the P is 0.000, which confirms its significance and validation.
TABLE 13.CFA Cost Furthermore, the fourth item: "Our company evaluates suppliers/subcontractors of PPP projects based on how well they meet the agreed budget," the factor loading is 0.554, with the P being 0.000 to confirm that this item is also validated and significant.The factor loading for item five: "Our company minimises the PPP project cost taking precedence over other objectives" stands at 0.603 with the P 0.000, which confirms the validation and significance of the first item.In addition, for the last item: "Overall PPP project cost performance was met based on baseline goals, targets, or expectations," the factor loading 0.659 and the P 0.000 show that it is both validated and significant.

FIGURE 9. CFA Cost
Table 14 shows the goodness of the Fit indexes of cost.The ratio is less than two, and the value at 1.792 indicates that the model is fit.The P-value with ratio of more than 0.05 and at value 0.064, with GFI more than 0.9 at the value of 0.976, TLI with ratio of more than 0.9 and the value of 0.954, and finally RMSEA's critical ratio is less than 0.08 and the value of 0.058 confirms this model to be significant and fit.Table 15 shows the Confirmatory Factor Analysis (CFA) of the four items of Time to Use.The result confirms that all four items are validated in Figure 10.The factor loading for item one, "Our company evaluates suppliers/subcontractors based on how well they meet schedules," is 0.658, with P being 0.000, which confirms the validation and significance of this item.The factor loading for item two: "Our Company minimises the PPP project cost taking precedence over other objectives" is 0.739 and the P being 0.000, again confirming that this item is validated and significant.Item three: "Our Company makes additional resources available to meet PPP project milestones and deadlines" have factor loading of 0.568, and the P is 0.000 making it validated and significant.Moreover, for the last item: "Our company takes corrective action to control progress against the PPP projects," the factor loading is 0.689, and the P is 0.000 confirming it to be validated and significant.Table 16 shows the goodness of the Fit indexes of time.The ratio is less than 2, and the value is 0.155, which shows that the model is fit.Furthermore, the P-value with a ratio more than 0.05 and value of 0.857, the GFI with ratio of more than 0.9 and value more than 0.999, TLI with a ratio of more than 0.9 and value of 1.024, and RMSEA with a ratio less than 0.08 and value of 0.000 confirm the model to be significant and fit.Table 17 shows the Confirmatory Factor Analysis (CFA) of the six items of procurement, out of twelve Figure 11.The result confirms that only six of the items are validated and significant.Item one: "Briefing is a process, which should have a clear goal and objectives" was deleted.Item two," An experienced person is needed to develop a brief."was removed.Item three, "A brief needs to make clear what the end-user requirements were eliminated.Item four, "During the briefing, the process of devising a brief need to be agreed by the key parties," was removed.Item five, "The public sector should lead throughout the briefing process, "was deleted.Furthermore, item six, "Briefing should be allocated with adequate time," was also deleted.The factor loading for item seven, "A consensus of the brief amongst the various" is 0.523, and the P is 0.000, which makes the item validated and significant.Item eight, "A consensus of the brief amongst the various stakeholders needs to be developed during the briefing process," has the factor loading 0.742 and the P is 0.000, which means that it is both validated and significant.The estimate for item nine: "The key parties should agree on the priority of decision to be made in briefing" is 0.803 at the significance level 0.000, which confirms it to be validated and significant.Item ten: "A schedule should be set for completing the brief" has an estimate of 0.824 and P equal to 0.000, which makes it validated and significant.Item eleven, "Flexibility in briefs should be provided to cater for possible changes," has an estimate of 0.816, therefore, approved, and P 0.000 makes it meaningful.Moreover, for the last item: "Decisions made should be recorded in detail," the evaluation at 0.848 and P at 0.000 show the validation and significance of the item.18 represents the goodness of the Fit indexes of procurement.The ratio being less than two and having the value is 0.467show that the model is fit.
The P-value with ratio of more than 0.05 and at 0.897 value, GFI more than 0.9 at the value 0.994, TLI with ratio more than 0.9, and the value 1. Table 21 represents the goodness of the Fit indexes of PPP Implementation.The ratio of less than two and the value 1.339 show that the model is fit.The P-value with a ratio of more than 0.05 and at 0.182 value, GFI more than 0.9 at the value 0.984, TLI with ratio more than 0.9 and value of 0.985, and finally RMSEA's critical ratio less than 0.08 and the value of 0.041 altogether confirm that the model of PPP Implementation is significant and fit in Figure 13.Table 22 also approved an extensive factor loading of PPP implementation measurements with all estimate values more than 0.5 and P-Value for 0.000.this study confirmed, there is six valid and significant measurement of PPP implementation in the setting of the construction project.A confirmatory evaluation of construct validity was offered for the measurement model, which defines and checks the relationships between the metric measures and their underlying constructs (Bentler, 1990).Only the direct causal interaction between the latent constructs, as posited by the theory (Anderson, 1988), was conducted.The confirmatory analysis of each dimension was also conducted to confirm the items.The next procedure was drawing the 2 nd order of the technology, organisational and project characteristics, which is the fundamental contribution of this study.To examine waiting time satisfaction is a mediating variable on the relationship between PPP success factors and PPP implementation, the indirect effect analysis was employed.The standardised factor loadings allow the researcher to arrange the order of entry of variables based on causal priority and is a useful tool for assessing interaction effects (Byrne B. M., 2001).This procedure enables partitioning of the unique variance explained by the interaction term above and beyond those accounted by the main effects.In this study, all hypothesised relationships were supported based on the SEM results.The path estimates for the hypothesis testing in the model show that all two hypothesised relationships were positively related to customer satisfaction (H1 and H2).The empirical results for each hypothesis achieved the objectives of this study.The hypothesis testing was accomplished by examining the completely standardised parameter estimates, the critical ratio, and probability level.The two-tailed test of significance was used to determine the importance of each path coefficient shown in Table 23.The result indicates that the direct relationship and indirect relationship hypotheses were consistent with expectations and statistically significant in the path direction.The findings of the latent constructs of exogenous and mediating variables of the model are significantly related to the customer satisfaction shown in Table 24.Explicitly, all hypotheses are is a confirmation of the mediating impact of stakeholder and procurement on the link between PPP implementation, Organisational and project characteristics, and technology.Hypothesis testing and path analysis reflect the total of the indirect and direct impact of variable interactions and highlight the goodness of model fit of the model that was hypothesised.As hypothesised, there is a definite link between PPP implementation, stakeholder and procurement with the results of SEM in evaluating the validity of empirical relationships between constructs of PPP success factors.The dimensions of technology, Organisational and project characteristics were positively related to stakeholder and procurement.The procurement and stakeholder were confirmed as a medium on the link between PPP's implementation and success factors via a path and direct-indirect analysis, acquisition and stakeholder.As such, the current study's aims were reached.The following chapter further explores the impact of the results on practise, study's limitations and ideas for future research.

FIGURE 1 .
FIGURE 1. Research framework and hypothesis

FIGURE 3 .
FIGURE 3. Flowchart of Covariance the Model

FIGURE
FIGURE 5. CFA Training TABLE 7. Goodness of Fit Index of CFA Training

TABLE 4 :
CFA Top Management Support FIGURE 4. CFA of Top Management Support

TABLE 5 .
Goodness of Fit Index of CFA Top Management CFA TRAINING (ORGANISATIONAL DIMENSIONS

Table 6
shows the Confirmatory Factor Analysis (CFA) of the Training items.All the factors have been validated and significant in Figure

TABLE 6
Table 7 highlights the goodness of the training suit indexes.The ratio is less than two, and the 2.230value indicates that the model is close appropriate.

TABLE 10
. CFA Usefulness FIGURE 7. CFA Model of Usefulness

TABLE 12
CFA COST (PROJECT CHARACTERISTIC DIMENSIONS)

TABLE 14 .
Goodness of Fit Index of CFA Cost

TABLE 16 .
Goodness of Fit Index of CFA Time

TABLE 21 .
Goodness of Fit Index of CFA Success of PPP Implementation