Project Management Institute

Structured equation modeling for determing ICT project success factors

Department of Applied Information Systems, Faculty of Management, University of Johannesburg, Auckland Park, Johannesburg, South Africa

Abstract

Projects, irrespective of type and industry, are initiated to provide value to the organization. This is also the case of ICT projects which have become an integral part of any organization. The problem is that ICT projects do not have a good track record, are failing most of the time, and do not provide value to the organization. This is an on-going phenomenon for over a decade. This article focuses on two aspects regarding ICT projects: the first is to investigate the factors that contribute to project success; and secondly, how these factors contribute individually as well as collectively to iCt project success. A scientific model that indicates how success factors influence project success has not been attempted before, and such a model adds to the current body of knowledge. A quantitative approach was followed in which survey respondents were requested to classify their ICT projects as successful, challenged, or failed. They also had to indicate how project outcome is measured as well as the factors that influence it. Structural equation modeling was done to determine if and how any single factor contributes to project outcome. The model revealed that stakeholder management, communication, and methodology are the driving factors behind ICT project success and that they are positively correlated to one another. The model can be applied by ICT project managers, thus ensuring a better project success rate. This model also opens the door for more debate regarding ICT project success and the manner in which various factors contribute to project success.

Keywords: information communication and technology; project success; South Africa; structural equation modeling

Introduction

It is the year 2002. Cooke-Davies (2002) ponders what factors contribute to project success. A decade or so later and there is still no answer and it seems as if information and communication technology (ICT) project success becomes the goal that everyone wants to achieve.

In the last few years, the project management discipline has matured through the publication of several standards (Ahlemann, Teuteberg, & Vogelsang, 2009; Crawford, Pollack, & England, 2006), best practices (Crawford & Pollack, 2007), research articles (Sewchurran, Smith, & Roode, 2010) and significant growth in its community of professional practitioners (Smith, Bruyns, & Evans, 2011). From this, it may be deduced that today, the majority of ICT projects should be completed successfully. However, literature suggests that this is not the case (Eveleens & Verhoef, 2010; Hyväri, 2006; Sheffield & Lemétayer, 2013). According to results from the various Chaos reports, only 35% of ICT projects are generally completed successfully, with the remainder of the projects being outright failures and being challenged. Comparative studies in South Africa provide the same results (Labuschagne & Marnewick, 2009; Marnewick, 2013; Sonnekus & Labuschagne, 2003). The ICT services’ market in South Africa has seen increased growth, with market research and advisory company IDC expecting the ICT services market to exceed US$14.59 billion (ZAR153.2 billion) in 2014. Based on the failure rate of 12% of ICT projects in South Africa, this means that US$1.75 billion (ZAR18.38 billion) could potentially be wasted on ICT projects in 2014. Given the current economic climate in South Africa as well as internationally, this is not an ideal situation.

ICT project success as well as the corresponding factors that contribute to ICT project success has been widely researched over the last decade or so. The one aspect that is missing from all of this research is how each individual factor contributes to the ICT project success dimensions. This article addresses this gap by means of structural equation modeling and indicates how each factor contributes to ICT project success. This model will guide project managers to focus on the various contributing factors and the factors’ interrelationship with each other ensuring ICT project success and ultimately the success of the organization.

The article is presented in four sections. The first section provides an overview of past and current literature shedding insight into project success and the factors that contribute to project success. The second section focuses on the research methodology and why structural equation modeling was used. The third section is on analysis and discusses the results and the impact of the model. The conclusion, which is the fourth section, provides insight into further research as well as possible limitations of the research and subsequent model.

Literature Review

The evolution of project management over the past few decades has brought about a paradigm shift regarding how successful projects are perceived and measured (Ika, 2009; Vadapalli & Mone, 2000). However, there is still a considerable amount of ambiguity regarding project success (Belassi & Tukel, 1996; Marnewick & Labuschagne, 2012). Furthermore, this is especially true for ICT projects, because these types of projects are becoming more prevalent within organizations across the globe and continue to fail at a disturbing rate.

Financial and Economic Impact

ICT solutions, delivered through projects and project management, do not only support everyday business operations but also enable the implementation of the organizational vision and mission (Cohen & Graham, 2001; Kendall & Rollins, 2003; Phillips, Bothell, & Snead, 2002). This proliferation of ICT has led to an increase in ICT spending in organizations (Gartner, 2013a; OECD, 2012; Ricknas, 2013). ICT-related projects cost approximately US$3,545 and US$3,588 billion globally in 2011 and 2012, respectively (Gartner, 2013b). It is expected that this amount will increase to US$4,188 billion by 2016. South Africa is not different. The IDC (2012) states that overall ICT spending reached US$12.91 billion (ZAR135.56 billion) in 2011 at an annual growth rate of 7.1%. However, there is a continuous element of concern. Eveleens and Verhoef (2010) revealed that since 24% of ICT projects were considered failures in 2009, this implies that US$777 billion were squandered in 2009 alone. Within the South African context, there is minimal difference once again. Research conducted by Marnewick (2012) reveals that 12% of ICT projects failed in 2011, which implies that US$1.55 billion (ZAR16.28 billion) was squandered in 2011. Furthermore, the IDC (2012) predicts that ICT spending in South Africa will increase to US$17.42 billion (ZAR182.91 billion) in 2016, implying that US$2.09 billion (ZAR21.95 billion) could potentially be squandered in 2016 unless the failure rate of ICT projects is addressed accordingly.

Given that significant investments are made on ICT-related projects, it is imperative that the performance rates of these projects are investigated in order to justify the significant investment required.

ICT Project Performance Rates

In order to gain a better understanding of the state of ICT projects, the Standish Group began to run a survey which later became known as the Chaos Chronicles (Standish Group, 1995). The purpose of the report is to document the success rates of ICT software development projects in particular (Standish Group, 1995; Yeo, 2002). However, since inception, the Chaos Chronicles have illustrated that little has changed regarding the success, challenge and failure rate of these types of projects.

Similar research has been conducted in South Africa which became known as the Prosperus report and focused on ICT projects in general (Sonnekus & Labuschagne, 2003). Figure 1 illustrates the Prosperus results from 2003 to 2011.

South African ICT Project Performance Rates (adapted from Marnewick, 2012)

Figure 1: South African ICT Project Performance Rates (adapted from Marnewick, 2012)

From 2003 to 2008, there was a disturbing trend that ICT projects were becoming more challenged as well as failing more at the expense of successful projects. This trend changed in 2011, however, because 59% of ICT projects were perceived as successful, signifying a 16% increase from the 2003 Prosperus report (Sonnekus & Labuschagne, 2003) and a 22% increase from the 2008 Prosperus report (Labuschagne & Marnewick, 2009). Apart from the successful projects, the remainder of the statistics are similar to those in the 2003 and 2008 reports. It is evident that there has been an average decline of 13% in failed projects from 2003 to 2008. Similarly, there has also been a steady decrease in challenged projects.

Project Success

Project success and project management success are often used interchangeably, but a multitude of researchers realized that it is necessary to make a clear distinction between the two (Baccarini, 1999; Cooke-Davies, 2002; de Wit, 1988; Ika, 2009; Van Der Westhuizen & Fitzgerald, 2005). Numerous researchers reiterate that true project success consists of two distinct components, viz. project management success and project product success (Baccarini, 1999; Bannerman, 2008; Camilleri, 2011; Cooke-Davies, 2002; de Wit, 1988; Ika, 2009; Jugdev, Perkins, Fortune, & White, 2013). These two components can be distinguished as follows (Baccarini, 1999; Bannerman, 2008; Erasmus & Marnewick, 2012; Pinkerton, 2003; Van Der Westhuizen & Fitzgerald, 2005):

  • Project management success primarily focuses on the processes applied to the project. This includes the successful achievement in terms of time, cost, and quality objectives. It is also imperative to consider the manner in which the project management process was conducted, as time, cost, and quality pertain to the efficiency of project execution.
  • Project product success focuses on the overall effect of the project’s final product or output. While project management success is distinguishable from product success, the successful outcomes of both of them are inseparably linked. “If the venture is not a success, neither is the project” (Pinkerton, 2003).

The following simplistic equation could arguably be formulated for project success (Erasmus & Marnewick, 2012; Van Der Westhuizen & Fitzgerald, 2005):

Project Success = Project Management Success + Project Product Success (1)

Project Management Success

Project management success consists of three fundamental components (Baccarini, 1999; Bannerman, 2008; Camilleri, 2011; Cooke-Davies, 2002):

  • Meeting time, cost, and quality objectives (project inputs and outputs);
  • Quality of the project management process (planning, controlling, and risk management);
  • Satisfying project stakeholders’ requirements and expectations where they pertain to the project management process.

Project Product Success

Project product success consists of three fundamental components (Baccarini, 1999; Bannerman, 2008; Camilleri, 2011; Cooke-Davies, 2002):

  • Project goal: meeting the project owner’s strategic organizational objectives;
  • Project purpose: ensuring the satisfaction of users’ requirements and expectations;
  • Project product: ensuring the satisfaction of stakeholders’ requirements and expectations where they pertain to the project product (final output).

Relationship Between Project Management Success and Project Product Success

Baccarini (1999) states that project management success is subordinate to project product success. The success criteria of time, cost, and performance of project management success have a lower rank when compared to the project product success criteria of project goal and purpose. This therefore explains, “why projects, which ought to be considered a disaster in project management terms, are perceived as successes simply because the higher-level objective was met” (de Wit, 1988). There is often disparity between satisfying the customer and meeting project performance objectives; many projects meet project specifications but fail to satisfy the customer and are thus classified as failures (Shenhar, Levy, & Dvir, 1997; Wateridge, 1998). Furthermore, Baccarini (1999) and Bannerman (2008) state that it would be foolish to believe that project management success does not influence the accomplishment of project product success. An effective project management process could influence and contribute towards project product success but is unlikely to prevent project product failure (Markus, Axline, Petrie, & Tanis, 2000; Young & Poon, 2013).

Critical Success Factors

Critical success factors (CSFs) refer to conditions, events and circumstances which contribute to the project result (Ika, 2009; Lim & Mohamed, 1999). Initial research on CSFs focused on the varying facets of project control, as these factors are considered the levers that a project manager could use to strengthen a project’s chances of success (Westerveld, 2003). The

majority of earlier research revolved around practitioners’ summarizing their experiences and was not based on scientific empiricism (Hyväri, 2006). Slevin and Pinto (1986) addressed the lack of scientific empiricism by developing the project implementation profile (PIP). The PIP consisted of ten crucial success factors: project mission, top management support, project schedule/plan, client consultation, personnel, technical tasks, client acceptance, monitoring and feedback, troubleshooting and communication. Furthermore, Pinto and Slevin (1988) augmented the PIP to include an additional four factors: characteristics of the project team leader, power and politics, environmental events and urgency. The concept behind the PIP was to establish a means for identifying the various facets which influence project success or failure (Hyväri, 2006).

Belassi and Tukel (1996) were among the first researchers to classify the various factors into more meaningful factor groups, that is, factors related to the project manager and project team members, the project, the organization, and the external environment. These factor groups are interrelated, because a factor from one group could impact a factor within another group (Belassi & Tukel, 1996). Moreover, a combination of factors from different groups could ultimately lead to project success or failure.

Sonnekus and Labuschagne (2003), Labuschagne and Marnewick (2009) as well as Marnewick (2012) conducted similar research but targeted South African ICT projects specifically. The previously mentioned research used the same fourteen CSFs to facilitate longitudinal analysis. The fourteen CSFs were requirements definition clarity, communication between team and customers, communication between project team members, business objectives clarity, understanding of user’s needs, project manager competency, executive support, handling of change, user involvement, change control processes, formal methodologies, user understanding of technology, auditing of processes, and support of innovative technology.

Given the ambiguity regarding the success of ICT projects, it is imperative to determine how these projects are perceived and measured. Moreover, it is imperative to identify which factors influence these projects as well as the inter-relationship between them. This research subsequently poses two research questions:

  1. How is ICT project success perceived and measured?
  2. Which factors influence ICT project success?

The next section deals with the implemented research methodology.

Research Methodology

There are two paradigms that could be employed when conducting research, quantitative and qualitative research (Biggam, 2011; Fox & Bayat, 2008; Thomas, 2003). The quantitative research paradigm focuses on answering the research question through the collection and statistical analysis of numerical data (Christensen, Johnson, & Turner, 2010; King, Keohane, & Verba, 1994). This paradigm collects data from specific instances and phenomena with the aim of providing results that are generalizable to other similar environments (Glesne & Peshkin, 1992; King, et al., 1994). Moreover, this paradigm acquires measurements and performs analyses which are easily repeatable by other researchers (King, et al., 1994).

Since the primary goal of this research is to determine how ICT project success is perceived and measured, as well as the factors which influence it, it is imperative that the obtained results were generalizable to a diverse array of organizations. The quantitative research paradigm was subsequently employed. This research applied a wide array of statistical analysis, and it was therefore imperative to express the data in numerical form. Statistical analysis facilitates deductive analysis, explanation, and prediction. Moreover, by understanding the significance of each factor, a prediction of whether an ICT project will succeed or fail should be possible.

The strategy implemented within this research was a survey in the form of a structured questionnaire as this allows the researcher to generalize the results (Denscombe, 2010; and Taylor, Sinha, & Ghoshal, 2006).

The unit of analysis or objects of study used in this research were project team members who had been involved in ICT projects specifically. ICT project team members were studied, as these individuals are involved throughout an ICT project’s lifecycle. This research employed probability sampling in the form of simple random sampling as this sampling technique has low bias and high generalizability (Biggam, 2011; Downing & Clark, 2010; Sekaran, 1992; and Walliman, 2005). The questionnaire was distributed via two means, viz. web-based as well as manual distribution.

Predictive modeling is a powerful technique to identify the influencing factors and their inter-relationship with regard to a particular phenomenon. This technique, however, is rarely used within the ICT project management domain.

Various researchers have explored the concept of predictive modeling within the field of project management (Boetticher, 2001; Cheng & Wu, 2008; de Barcelos Tronto, da Silva, & Sant’Anna, 2008; Finnie, Wittig, & Desharnais, 1997; Gray & MacDonell, 1997; Karunanithi, Whitley, & Malaiya, 1992; Kim, Han, Kim, & Park, 2009; Ling & Liu, 2004; Wang & Yu, 2011). Classical statistical techniques have often been employed to develop predictive models (Razi & Athappilly, 2005). Linear, multiple and logistic regression are arguably the three most prevalent statistical techniques employed within the project management domain (Berlin, Raz, Glezer, & Zviran, 2009; de Barcelos Tronto, et al., 2008; Kim, et al., 2009; Ling & Liu, 2004; Wang & Yu, 2011). However, many researchers began to explore machine learning as an alternative tool for predictive modeling in an attempt to overcome the inherent limitations of regression (de Barcelos Tronto, et al., 2008; Lee & Jung, 2000; Razi & Athappilly, 2005; Shang, Lin, & Goetz, 2000; West, Brockett, & Golden, 1997). The most prevalent machine learning technique employed within project management is artificial neural networks (ANN) (de Barcelos Tronto, et al., 2008; Gray & MacDonell, 1997; Ling & Liu, 2004; Zhang, Keil, Rai, & Mann, 2003). However, although ANN provides better results than regression techniques, it requires a significant amount of time and money to implement and thus is not viewed as resource efficient. Researchers subsequently turned to the multivariate technique of structural equation modeling (SEM) (Kim, et al., 2009). SEM provides a more perceptive understanding of the interrelationships between the various variables compared with regression and ANN, as it can measure direct and indirect influences and impacts between different latent and observed variables (Kim, et al., 2009). SEM also allows the complex relationships between variables to be viewed visually via graphical representation (Doloi, Iyer, & Sawhney, 2011; Foster, Barkus, & Yavorsky, 2006; Kim, et al., 2009). SEM has subsequently been selected as the predictive modeling technique for this research.

Results and Discussion

A total of 1731 responses were received and 4330 ICT projects were assessed. It is important to note that this research focused on ICT projects in general.

Project Performance Rates

Respondents were asked to indicate how many projects failed, how many were challenged, and how many were successful. Figure 2 illustrates project success rates over the past decade and compares the 2013 results with those from the 2003 and 2008 Prosperus reports as well as from research conducted by Marnewick (2012). The trend pattern in Figure 2 implies that the percentage of successful projects has decreased by 25% when compared with the 2011 results. Furthermore, failed projects have increased by a significant 20% from 2011. The increase in failed projects could be at the expense of successful projects. It is interesting to note that challenged projects exhibit a consistent trend pattern as these results have not fluctuated as much as the failed and successful project results. Challenged projects have stayed within a band of 29% and 36% over the past decade. These figures could be considered an ominous sign for the future of ICT projects within South Africa.

South African ICT project performance over the past decade (2003-2013)

Figure 2: South African ICT project performance over the past decade (2003-2013)

Project Success Criteria

The first research question was to determine how ICT project success is perceived and measured. The respondents were asked to indicate which success criteria are the most important when measuring ICT project success. It must be noted that the 2003 and 2008 Prosperus reports did not focus on determining project success criteria. Table 1 illustrates the project success criteria results for 2011 and 2013.

Table 1: Project success criteria comparison (2011 – 2013)

Ranking Success Criteria 2011 (%) 2013 (%) Difference (%) Change
1 Quadruple constraint 73 85 12 arrow
2 User satisfaction 72 84 12 arrow
3 Met business objectives 52 82 30 arrow
4 Met project requirements 70 81 11 arrow
5 Delivery of business benefits 68 80 12 arrow
6 Stakeholder satisfaction 64 77 13 arrow
7 Sponsor satisfaction 60 72 12 arrow
8 System implementation 57 71 14 arrow
9 System use 52 65 13 arrow
10 Triple constraint 46 60 14 arrow
11 Steering group satisfaction 43 59 16 arrow

Quadruple constraint and user satisfaction were the two most highly ranked project success criteria with 85% and 84%, respectively. There was, however, one particular success criterion which rose significantly compared with 2011’s research results, that is, met business objectives. This success criterion was ranked third last in 2011 with a mere 52%, but now ranked third overall with a significant 82%. This is an astonishing 30% increase over two years. It could be argued that this is in line with the view that there has been a paradigm shift in the way project success is perceived or measured, because organizations no longer rely on the triple constraint to determine project success, but rather assess project success based on the quadruple constraint, user satisfaction and whether business objectives were met. Furthermore, many organizations have been struggling to recover from the financial crisis of 2008 and may thus have paid more attention to the ability of ICT solutions to help them recover. This in turn has forced organizations to pay more attention to ICT projects and ensure that these projects meet the objectives of the business and facilitate the recovery process.

Project-Influencing Factors

The second research question was to identify the factors which influence ICT project success. The 2003 and 2008 Prosperus reports identified fourteen key factors that could influence the outcome of an ICT project. The same fourteen factors were used within this research. The respondents were asked to indicate which factors influence ICT project success through Likert scale questions. Table 2 lists the success factors ranked as “very important” for 2003, 2008, 2011, and 2013. Furthermore, the factors are sorted based on the response percentages, while the top three factors have been highlighted.

Table 2: Project success factors comparison (2003 – 2013)

Ranking Factors influencing successful projects 2003 2008 2011 2013
1 Requirements definition clarity 4 6 6 1
2 Communication between team and customers 3 2 1 2
3 Communication between project team members 1 10 10 3
4 Business objectives clarity 6 5 5 4
5 Understanding of user’s needs 2 9 9 5
6 Project manager competency 5 12 12 6
7 Executive support 8 1 2 7
8 Handling of change 9 3 3 8
9 User involvement 7 4 4 9
10 Change control processes 9 8 8 10
11 Formal methodologies 11 11 11 11
12 User understanding of technology 12 7 7 12
13 Auditing of processes 14 14 14 13
14 Support of innovative technology 13 13 13 14

The results illustrate how the various success factors have changed in rank over the past decade. It is evident that there has been a shift from 2003 to 2011. The 2008 and 2011 results exhibited the same factors but the rankings changed. The latest results, 2013, indicate that requirements definition clarity is the main influencing factor of ICT project success while communication between team and customers is ranked second and communication between project team members is ranked third. It is, however, important to note that there is one highly ranked factor shared over the past decade, that is, communication between team and customers. The results highlight that the factors are subjective and therefore require that a scientific model be developed to predict ICT project success.

Predictive Model for ICT Project Success

The construction of the predictive model included four key phases, that is, data preparation, developing a theoretical model, performing exploratory factor analysis (EFA) as well as confirmatory factor analysis (CFA).

Data preparation. Data preparation involves data editing, data codification, blank response handling, and data file preparation. Data editing involved deleting all nonessential data from the initial dataset, that is, only responses pertaining to the fourteen project success influencing factors were retained. These success factors were given distinct observed variable names as depicted in Table 3. It is important to note that, from here on, the fourteen project success factors will be referred to by their observed variable names and that the latent variables identified during the EFA phase will be referred as factors.

Table 3: Project Success Factor Naming Codification

Observed Variable Name Project Success Factor
b20.1 Adequate handling of change
b20.2 Good communication between team and customers
b20.3 Good communication between project team members
b20.4 Adequate project manager competency
b20.5 Maximum support of innovative technology
b20.6 Adequate user understanding of technology
b20.7 Positive executive support
b20.8 Clear business objectives
b20.9 Good understanding of user’s needs
b20.10 Clear requirements definition
b20.11 Frequent user involvement
b20.12 Adequate change control processes
b20.13 Appropriate formal methodologies
b20.14 Correct auditing of processes

A Likert scale was used for the responses pertaining to the fourteen project success influencing factors. These responses were codified as follows: not at all important (1), slightly important (2), neutral (3), somewhat important (4), very important (5), not sure (6). A two-pronged approach was employed when handling blank responses. Firstly, 356 of the 1731 responses were deleted as more than 25% of the values were missing from these responses. Secondly, the 1375 remaining responses were assigned the value “6” for missing values as “6” represents “not sure” as per the codification standard.

Development of a theoretical model. Prior to constructing the predictive model for ICT project success, a theoretical model was initially developed based on the literature review. Figure 3 visually maps the fourteen factors to each of the six components of ICT project success, that is, triple constraint, quality of project management process, stakeholder satisfaction pertaining to the project management process, project goal, project purpose, and stakeholder satisfaction pertaining to the project product or output.

Theoretical Project Success Model

Figure 3: Theoretical Project Success Model

Primary exploratory factor analysis. The purpose of EFA is to assess the dimensionality of the observed variables attained from the original questionnaire, for example, EFA statistically assesses the observed variables and condenses them into fewer latent variables that are simpler to comprehend (Brown, 2006; Field, 2009).

EFA validity is assessed using a number of adequacy measures as well as convergent validity measures. The first test for adequacy is the Kaiser-Meyer-Olkin (KMO) measure (Bryde, 2008; Field, 2009; Gaskin, 2013a; Lee & Yu, 2012). Researchers and practitioners consider a KMO value greater than 0.9 as “marvelous” (Field, 2009; Gaskin, 2013a, 2013b; Hutcheson & Sofroniou, 1999; Kaiser, 1974). The KMO value was 0.93 and is therefore considered as an acceptable value for EFA. Furthermore, this value implies that the EFA results are valid.

The second test for adequacy involves assessing the extraction values within the communalities table. The extraction values are shown in Table 4. Gaskin (2013b) states that these values should be greater than 0.3. All the extraction values are greater than 0.3, implying that the second test for adequacy is valid.

Table 4: Extraction Values

Communalities
Initial Extraction
b20.1 0.751 0.788
b20.2 0.733 0.786
b20.3 0.737 0.803
b20.4 0.799 0.844
b20.5 0.791 0.823
b20.6 0.576 0.585
b20.7 0.761 0.795
b20.8 0.600 0.590
b20.9 0.699 0.700
b20.10 0.664 0.647
b20.11 0.758 0.797
b20.12 0.651 0.703
b20.13 0.672 0.760
b20.14 0.619 0.696
      Extraction method: Maximum likelihood

The third test for adequacy focuses on assessing the total variance explained. Table 5 presents the explained total variance results. The EFA identified three factors. The cumulative percentage was 73.69%, implying that the three identified factors account for 73.69% of the total variance within the dataset. This result was deemed acceptable as nearly 75% of the variance was explained by the three identified factors (Gaskin, 2013b).

Table 5: Total Variance Explained

Factor Total Percent of Variance Cumulative %
1 6.612 47.226 47.226
2 3.131 22.361 69.587
3 0.574 4.103 73.690
      Extraction method: Maximum likelihood

Goodness-of-fit was assessed as the fourth test for adequacy. The goodness-of-fit results are given in Table 6. The significance value is less than 0.05, implying that the EFA results are valid and adequate (Gaskin, 2013b).

Table 6: Goodness-of-fit results

Goodness-of-fit Test
Chi-Square df Significance
530.278 52 0.000

The fifth and final test for adequacy was to assess the nonredundant residuals of the reproduction correlations table. It is recommended that this value is less than 5% as a higher value could potentially distort the CFA results as well as the final SEM (Gaskin, 2013b). The nonredundant residuals for this research were 4%, implying that the EFA results are valid and adequate.

EFA validity is also assessed through convergent validity. Convergent validity is determined by assessing the factor loadings within the pattern matrix (Gaskin, 2013b). The EFA pattern matrix is depicted in Table 7. The observed variable factor loadings have been highlighted to show which variables are associated with which factor. Gaskin (2013b) and Hair, Black, Babin, Anderson, and Tatham (2006) are of the opinion that the loadings should be greater than 0.5 when performing SEM and that the average loading within each factor should be greater than 0.7. The results indicate that each factor loading is well above 0.5 and that the average loadings are above 0.7. This implies that convergent validity has been achieved.

Table 7: Pattern Matrix

Pattern Matrix
Observed Variable Factor
1 2 3
b20.1 0.890 0-.001 -0.004
b20.2 -0.020 0.949 -0.080
b20.3 0.011 0.976 -0.119
b20.4 0.928 0.061 -0.066
b20.5 0.893 -0.076 0.080
b20.6 0.061 0.219 0.562
b20.7 0.892 0.027 -0.019
b20.8 -0.020 0.653 0.155
b20.9 -0.035 0.703 0.185
b20.10 0.061 0.671 0.144
b20.11 0.891 -0.015 0.014
b20.12 -0.020 0.214 0.677
b20.13 -0.008 0.046 0.841
b20.14 -0.002 0.012 0.827
Average loadings 0.899 0.790 0.727

      Extraction method: Maximum likelihood
      Rotation method: Promax with Kaiser normalization4

      ARotation converged in five iterations

Field (2009) and Graham, Guthrie, and Thompson (2003) caution that the structure matrix should be assessed as well to ensure convergent validity. The EFA structure matrix is depicted in Table 8, and the loadings have been highlighted once again. The structure matrix results cross validate the pattern matrix results and thus further validate the EFA convergent validity.

Table 8: Structure Matrix

Structure Matrix
Observed Variable Factor
1 2 3
b20.1 0.888 0.257 0.373
b20.2 0.224 0.884 0.607
b20.3 0.247 0.892 0.601
b20.4 0.918 0.285 0.373
b20.5 0.905 0.245 0.404
b20.6 0.364 0.649 0.748
b20.7 0.892 0.274 0.379
b20.8 0.238 0.761 0.626
b20.9 0.249 0.828 0.685
b20.10 0.319 0.794 0.661
b20.11 0.893 0.257 0.382
b20.12 0.330 0.704 0.825
b20.13 0.363 0.660 0.871
b20.14 0.352 0.617 0.834
      Extraction method: maximum likelihood
      Rotation method: Promax with Kaiser normalization

In order to ensure EFA reliability, Cronbach’s alpha reliability test was applied. The Cronbach’s alpha threshold was set at greater than 0.7 (Doloi, et al., 2011; Field, 2009; Gaskin, 2013c; Roh, Ahn, & Han, 2005; Sohn, Kim, & Moon, 2007; Sohn & Moon, 2003). Table 9 illustrates the reliability results. The Cronbach’s alpha results indicate that all factor groupings were above the 0.7 threshold. Factor grouping 1 was 0.953 while factor groupings 2 and 3 were 0.92 and 0.892, respectively. These results confirm the reliability of the EFA.

Table 9: Cronbach’s Alpha Results

Factor groupings Cronbach’s alpha result ( > 0.7 threshold)
Factor 1: b20.1, b20.4, b20.5, b20.7 and b20.11 0.953
Factor 2: b20.2, b20.3, b20.8, b20.9 and b20.10 0.92
Factor 3: b20.6, b20.12, b20.13 and b20.14 0.892

Figure 4 illustrates the EFA results using the factor plot in the rotated factor space diagram generated by SPSS. Furthermore, the observed variables are grouped to illustrate the three identified factors. The EFA results indicate that variables b20.1, b20.4, b20.5, b20.7 and b20.11 are associated with factor 1 or latent variable 1 (solid line grouping). Furthermore, variables b20.2, b20.3, b20.8, b20.9 and b20.10 are associated with factor 2 (dotted line grouping) while b20.6, b20.12, b20.13 and b20.14 are associated with factor 3 (dashed line grouping). The identified factors are named by mapping the observed variables to the identified factors.

EFA Factor Plot Diagram

Figure 4: EFA Factor Plot Diagram

Factor 1 was identified as stakeholder management as the observed variables pertain predominantly to issues related to the various stakeholders. For example, stakeholder change requests are inevitable during any type of project and especially for ICT projects. Adequate handling of change (b20.1) pertains particularly to stakeholder management as the handling of these change requests is often paramount to satisfying the stakeholders. The project manager’s competency (b20.4) also plays a pivotal role when addressing stakeholder management. For example, a competent project manager should not only be knowledgeable about the various project management principles, but should also have the appropriate soft skills or people skills to manage the various stakeholders involved effectively. It could be argued that the project team members are one of the key stakeholders in any ICT project because they are fundamental to delivering the requested ICT solution. It is therefore imperative that the project team have maximum support of innovative technology (b20.5). These tools would facilitate an effective and efficient project management process. ICT projects, like any other type of project, require sufficient resources to be allocated to a project throughout its life cycle and this often achieved through positive executive support (b20.7). Successful stakeholder management therefore includes managing the support of senior and executive management, who are often responsible for allocating the required resources for any given project. Users are another key stakeholder within ICT projects since they ultimately use the requested ICT solution for their day-to-day tasks. Frequent user involvement (b20.11) is therefore a fundamental component of effective stakeholder management; if users are neglected, the ICT solution delivered will not meet the expectations or the needs of those who are expected to use it.

Factor 2 was identified as communication, since the observed variables focus on issues related specifically to communication and collaboration between project stakeholders. Effective communication is often touted as a key element of project success; thus good communication between team and customers (b20.2) is a prerequisite for project success. It is therefore essential that the team fully understand the customers’ as well as the users’ needs by interacting with them on a continuous basis. Furthermore, each team member has a specific role and responsibility during a project. Therefore, it is paramount to have good communication between project team members (b20.3) to ensure that each team member understands what is required of them if the project is to be successful. Good communication between team and customers (b20.2) and good communication between project team members (b20.3) also facilitate the formation of clear business objectives (b20.8), clear requirements definition (b20.10) and good understanding of user’s needs (b20.9). For example, poor communication could lead to the project team members and various stakeholders misunderstanding one another, resulting in the poor articulation of business and user needs. Furthermore, if team members have poor rapport with the customer, this will inevitably lead to an ambiguous and blurred requirements definition, as the team members will not fully understand the customers’ and users’ needs.

The third factor was identified as methodology, as the observed variables primarily pertain to project management practices and principles. For example, the most evident was appropriate formal methodologies (b20.13), which specifically focuses on the use of a formal project management methodology during an ICT project. Furthermore, a formal methodology should include adequate change control processes (b20.12) which not only facilitate effective change management, but also ensure that stakeholder satisfaction is maintained throughout the project life cycle. A formal methodology should also include practices and principles which facilitate the correct auditing of processes (b20.14). For example, the project management process has many subprocesses that should be audited to ensure that they are correctly executed. If a process is not executed correctly, it could have an adverse effect on ultimate project success and lead to, among other things, poor communication and stakeholder management. Adequate user understanding of technology (b20.6) was considered an anomaly within the third factor, as it could be argued that this observed variable relates more to stakeholder management. However, while this was considered an anomaly, the primary EFA results were preserved to ensure that the SEM process was not compromised.

Secondary exploratory factor analysis results. In order to ascertain whether there were additional factors present, a secondary EFA was performed. The secondary EFA employed a different strategy compared to the primary EFA. Table 10 shows how the observed variables were analyzed.

Table 10: Secondary EFA tests

Secondary EFA iteration Observed variables analysed
1 b20.1, b20.4, b20.5, b20.7 and b20.11 as well as b20.2, b20.3, b20.8, b20.9 and b20.10
2 b20.1, b20.4, b20.5, b20.7 and b20.11 as well as b20.6, b20.12, b20.13 and b20.14
3 b20.2, b20.3, b20.8, b20.9 and b20.10 as well as b20.6, b20.12, b20.13 and b20.14
4 b20.1, b20.4, b20.5, b20.7 and b20.11
5 b20.2, b20.3, b20.8, b20.9 and b20.10
6 b20.6, b20.12, b20.13 and b20.14

The first iteration confirmed that variables b20.1, b20.4, b20.5, b20.7 and b20.11 belong to factor 1 while variables b20.2, b20.3, b20.8, b20.9 and b20.10 belong to factor 2. The second iteration confirmed that variables b20.1, b20.4, b20.5, b20.7 and b20.11 belong to factor 1 while variables b20.6, b20.12, b20.13 and b20.14 belong to factor 3. The third iteration confirmed that variables b20.2, b20.3, b20.8, b20.9 and b20.10 belong to factor 2 while variables b20.6, b20.12, b20.13 and b20.14 belong to factor 3. The fourth, fifth and sixth iteration confirmed the factor groupings as SPSS could not reduce the variables into further factors.

Confirmatory factor analysis. CFA builds on the EFA results. This phase applies the pattern matrix (Table 7) generated during the primary EFA. The pattern matrix was imported into AMOS and the measurement model was subsequently generated (Figure 5). The measurement model implies that latent variables stakeholder management, communication and methodology are correlated to some degree. Observed variables b20.1, b20.4, b20.5, b20.7 and b20.11 are predictors of the stakeholder management latent variable. Furthermore, observed variables b20.2, b20.3, b20.8, b20.9 and b20.10 are predictors of the communication latent variable, while b20.6, b20.12, b20.13 and b20.14 are predictors of the methodology latent variable. In order to ensure that the overall model fit was not compromised, an error term was included to make provision for the potential error of a predicting variable (Anglim, 2007). The error terms are denoted by e1–e14.

The Measurement Model

Figure 5: The Measurement Model

Figure 6 illustrates the initial structural model generated after the estimation calculations were executed. The initial structural model indicates that observed variables b20.1, b20.4, b20.5, b20.7 and b20.11 are very strong predictors of stakeholder management because their regression weightings were 0.89, 0.92, 0.90, 0.89 and 0.89, respectively. Furthermore, observed variables b20.2, b20.3, b20.8, b20.9 and b20.10 are very strong predictors of communication as their regression weightings were 0.85, 0.85, 0.79, 0.86 and 0.83, respectively. With regard to the methodology latent variable, observed variables b20.6, b20.12, b20.13 and b20.14 are very strong predictors and their weightings were 0.77, 0.85, 0.86 and 0.82, respectively. Figure 6 also illustrates that stakeholder management and communication have a positive correlation or covariance of 0.3, while stakeholder management and methodology have a positive correlation of 0.42. Furthermore, communication and methodology have a positive correlation of 0.81.

Initial Structural Model

Figure 6: Initial Structural Model

In order to assess the model fit, a number of model fit measure were applied. Table 11 lists the model fit measures applied as well as the cut-off levels for this research and the results of the initial model fit measures.

Table 11: Model fit measures, cut-off levels and initial results

Model Fit Measures Cut-Off Levels Employed Results Reference
Absolute fit measures CMIN/DF (Chi-squared / Degrees of freedom) ≤ 5 9.727 Gaskin (2013d);Marsh and Hocevar (1985); Ullman (1996); McKinney, Yooon, & Zahedi (2002); Roh et al. (2005); Yatim (2008);
RMR (Root Mean Square Residual) ≤ 0.05 0.063 Roh et al. (2005); Tabachnick and Fidell (1996)
GFI (Goodness-of-Fit Index) ≥ 0.9 0.926 Doloi et al. (2011); Kim et al. (2009); Roh et al. (2005);Tabachnick and Fidell (1996)
Relative fit measures NFI (Normal Fit Index) ≥ 0.9 0.958 Doloi et al. (2011); Stahl (2008);Tabachnick and Fidell (1996); Yatim (2008)
TLI (Tucker-Lewis Index) ≥ 0.9 0.953 Doloi et al. (2011); Hair et al. (2006); Stahl (2008); Yatim (2008)
CFI (Comparative Fit Index) ≥ 0.95 0.962 Anglim (2007); Doloi et al. (2011); Gaskin (2013d); Hair et al. (2006); Roh et al. (2005); Stahl (2008); Tabachnick and Fidell (1996); Yatim (2008);
Fit measures based on the non-central chi-square distributions RMSEA (Root Mean Square Error of Approximation) ≤ 0.05 0.080 Anglim (2007); Doloi et al. (2011); Gaskin (2013d); Hair et al. (2006); Kim et al. (2009); Stahl (2008); Yatim (2008)
PCLOSE (RMSEA significance) > 0.05 0.000 Gaskin (2013d)

These results indicate that the GFI, NFI, TLI and CFI adhered to the predetermined cut-off levels in Table 11, while the CMIN/DF, RMR and RMSEA did not. The results imply that the initial structural model could not be accepted as not all of the model fit measures were acceptable.

These results required modification of the model to improve the model fit measures. The first modification made was to discard the observed variable with the lowest regression weighting, that is, b20.6 (0.77). This omission was deemed valid as observed variable b20.6 (adequate user understanding of technology) was considered an anomaly within the methodology factor group. Table 12 reveals the second set of model fit measures once the estimates were recalculated. The second set of model fit measures indicate that while CMIN/DF, RMR, RMSEA and PCLOSE all moved closer to the acceptable cut-off level, these measures were still not at an acceptable level. On the other hand, the GFI, NFI, TLI and CFI model fit measures all improved slightly. The overall fit of the model was therefore positively affected by the omission of observed variable b20.6 (adequate user understanding of technology).

Table 12: Second set of model fit measures

Model Fit Measures Cut-Off Levels Employed Results Change
Absolute fit measures CMIN/DF (Chi-squared / Degrees of freedom) ≤ 5 9.174 darrow
RMR (Root Mean Square Residual) ≤ 0.05 0.055 darrow
GFI (Goodness-of-Fit Index) ≥ 0.9 0.934 arrow
Relative fit measures NFI (Normal Fit Index) ≥ 0.9 0.964 arrow
TLI (Tucker-Lewis Index) ≥ 0.9 0.960 arrow
CFI (Comparative Fit Index) ≥ 0.95 0.968 arrow
Fit measures based on the non-central chi-square distributions RMSEA (Root Mean Square Error of Approximation) ≤ 0.05 0.077 darrow
PCLOSE (RMSEA significance) > 0.05 0.000 =

Further model modification was facilitated through the use of the modification indices generated by AMOS. Modification indices inform one of the underlying relationships which have not yet been defined in the structural model. The relationship needs to be indicated in the model by adding a covariance or correlation based on the modification indices. When dealing with modification indices, it is suggested that they be placed in descending order starting with the largest index (Brown, 2006; Gaskin, 2013d; Jöreskog, 1993). However, Brown (2006) argues that a covariance should not be added between an error term and a latent variable as it does not make logical sense. The largest modification index was 301.253 between e6 and e7. Table 13 gives the model fit measures after the covariance between e6 and e7 had been included and estimates recalculated.

Table 13: Final and accepted model fit measures

Model Fit Measures Cut-Off Levels Employed Results Change
Absolute fit measures CMIN/DF (Chi-squared / Degrees of freedom) ≤ 5 3.227 darrow
RMR (Root Mean Square Residual) ≤ 0.05 0.043 darrow
GFI (Goodness-of-Fit Index) ≥ 0.9 0.978 arrow
Relative fit measures NFI (Normal Fit Index) ≥ 0.9 0.988 arrow
TLI (Tucker-Lewis Index) ≥ 0.9 0.989 arrow
CFI (Comparative Fit Index) ≥ 0.95 0.991 arrow
Fit measures based on the non-central chi-square distributions RMSEA (Root Mean Square Error of Approximation) ≤ 0.05 0.040 darrow
PCLOSE (RMSEA significance) > 0.05 0.995 arrow

The results in Table 13 imply that CMIN/DF, RMR, RMSEA and PCLOSE all improved substantially and to an acceptable level. Furthermore, the GFI, NFI, TLI and CFI once again improved and were much closer to the perfect model fit level of 1. Based on the above model fit results, the structural model illustrated in Figure 7 was accepted as the final model for this research.

Final structural model

Figure 7: Final structural model

Figure 7 indicates that the regression weightings for observed variables b20.1, b20.4, b20.5, b20.7 and b20.11 remained at 0.89, 0.92, 0.90, 0.89 and 0.89, respectively. The regression weightings for observed variables b20.2, b20.3, b20.8, b20.9 and b20.10 changed to 0.79, 0.79, 0.81, 0.88 and 0.85, respectively. Furthermore, the regression weightings for observed variables b20.12, b20.13 and b20.14 changed to 0.85, 0.87 and 0.82, respectively. Although there were changes in the regression weightings, the weightings maintained their very strong predictor status. The correlation between stakeholder management and communication changed to 0.31. The correlation between stakeholder management and methodology remained at 0.41, while the correlation between communication and methodology remained at 0.80. Furthermore, there was an addition to the final structural model, namely the correlation or covariance between error term e6 and e7. AMOS calculated the covariance as 0.55, indicating that there was a strong positive correlation between the error terms.

A number of implications are indicated within the final structural model:

  • The first implication is that adequate handling of change (b20.1), adequate project manager competency (b20.4), maximum support of innovative technology (b20.5), positive executive support (b20.7) and frequent user involvement (b20.11) are very strong predictors of successful stakeholder management as their regression weightings are 0.89, 0.92, 0.90, 0.89 and 0.89, respectively. The stakeholder management latent variable corresponds to the views within project management standards such as P2M, A Guide to the Project Management Body of Knowledge (PMBOK® Guide) – Fifth Edition and ISO 21500 which reiterate the importance of managing stakeholders effectively to ensure overall project success (International Organization for Standardization, 2012; Ohara, 2005; Project Management Institute, 2013). For example, P2M has a chapter dedicated to relationship management which includes practical guidelines for successfully managing project stakeholders (Ohara, 2005). Comparably, ISO 21500 has a dedicated section named “Stakeholders and Project Organization, which discusses the importance of identifying project stakeholders as well as their roles and responsibilities (International Organization for Standardization, 2012). Furthermore, the fifth version of the PMBOK® Guide now includes a tenth Knowledge Area on Stakeholder Management. This new Knowledge Area focuses on identifying project stakeholders, planning stakeholder management, managing stakeholder engagement and controlling stakeholder engagement (Project Management Institute, 2013). It could therefore be stated that the final model not only empirically validates the importance of managing stakeholders throughout a project’s life cycle, but also empirically validates the inclusion of Stakeholder Management as a new Knowledge Area within the PMBOK® Guide.
  • The second implication is that good communication between team and customers (b20.2), good communication between project team members (b20.3), clear business objectives (b20.8), good understanding of user’s needs (b20.9) and clear requirements definition (b20.10) are very strong predictors of effective communication, as their regression weightings are 0.79, 0.79, 0.81, 0.88 and 0.85, respectively. Similar to stakeholder management, communication is an important topic covered in project management standards such as P2M, PMBOK® Guide and ISO 21500. For example, P2M has a chapter dedicated to project communications management which includes practical guidelines for ensuring effective communication between the various stakeholders (Ohara, 2005). Similarly, ISO 21500 includes sections such as Plan Communications, Distribute Information and Management Communications as subprocesses within the greater planning, implementing, and controlling project process groups. Furthermore, Project Communications Management is one of the ten Knowledge Areas within the PMBOK® Guide, and it focuses on planning communications management, managing communications, and controlling communications. Numerous researchers and practitioners have documented the influence and impact of communication on overall project success (Camilleri, 2011; Hyväri, 2006; Labuschagne & Marnewick, 2009; Marnewick, 2012; Slevin & Pinto, 1986; Sonnekus & Labuschagne, 2003). It could therefore be stated that the final model empirically validates the importance of effective communication on overall project success.
  • The third implication is that adequate change control processes (b20.12), appropriate formal methodologies (b20.13) and correct auditing of processes (b20.14) are very strong predictors of a comprehensive and effective methodology since their regression weightings are 0.85, 0.87 and 0.82, respectively. The Project Management Institute (2013) defines a methodology as “a system of practices, techniques, procedures, and rules used by those who work in a discipline.” The change control and auditing processes are pivotal components of any project management methodology. Furthermore, the model suggests that project success is significantly influenced by the implementation of a comprehensive methodology. It could be argued that project failure is inevitable if a comprehensive methodology is not implemented as this would lead to an incoherent, disorganised project where none of the stakeholders know their roles and responsibilities or what is expected of them and when. This model therefore empirically validates the importance of adopting and implementing a methodology throughout a project’s life cycle since a project would inevitably fail if no comprehensive methodology is implemented.
  • The fourth implication is that communication has an unquestionable effect on stakeholder management and vice versa. This relationship could be seen as logical since effective stakeholder management can only occur if there are effective channels of communication present throughout a project’s life cycle. Furthermore, it could be argued that communication is facilitated by means of effective stakeholder management practices and procedures, which dictate how the various stakeholders communicate and collaborate together.
  • The fifth implication is that there is a relationship between stakeholder management and methodology as the correlation coefficient is 0.41. This implies that there is a fairly strong positive relationship between the two and that they are directly influenced by each other. It could be argued that stakeholder management is dependent and facilitated by the adoption and implementation of a project management methodology, since a comprehensive methodology governs and dictates how stakeholders should be managed for a project to be successful. Furthermore, an unclear and vague methodology could lead to poor stakeholder management because it would probably lack well-defined practices and procedures for governing the management of stakeholders.
  • The sixth implication is that there is a relationship between communication and methodology as the correlation coefficient is 0.80. This positive correlation is very strong and implies that communication and methodology are directly influenced by each other. This positive relationship could be seen as logical as a methodology should provide communication guidelines by documenting practices and procedures for communicating with the various stakeholders. Furthermore, similar to the fifth implication, if a methodology lacks communication guidelines, it is inevitable that poor, ineffective communication will take place.
  • The seventh implication is that the final structural model is simpler to understand and thus more interpretable since it refines and simplifies the project success concepts within the theoretical model. For example, it is apparent in Figure 3 that each factor is associated with more than one latent or underlying variable, while Figure 7 illustrates that each factor belongs to only one latent variable. Furthermore, the final model implies that adequate user understanding of technology (b20.6) is not necessary for ICT project success because the omission of this observed variable improved the validity and reliability of the final structural model.
  • The eighth implication is that the model confirms literature as it illustrates that ICT project success is dependent on project management success as well as project product success. For example, methodology is directly related to project management process quality, as it should dictate and govern how a project is executed, especially with regard to the triple constraint. Furthermore, stakeholder management and communication are key components of a quality project management process, since they play a pivotal role in ensuring stakeholder satisfaction as well as facilitating the realisation of the triple constraint. On the other hand, stakeholder management and communication are also vital to realizing the project goal and purpose as well as ultimate stakeholder satisfaction. For example, effective stakeholder management and communication ensure that the project goal and purpose are accurately articulated, which in turn ensures ultimate stakeholder satisfaction. Furthermore, similar to project management success, a comprehensive methodology should dictate and govern how the project goal and purpose are defined to ensure benefits realisation and ultimate stakeholder satisfaction.

Conclusion

The financial and economic impact of failed ICT projects is profound. Research states that the performance rate of ICT projects continues to fluctuate and these projects continue to fail at a disturbing rate.

There is much ambiguity regarding project success. However, numerous researchers reiterate that true project success consists of two distinct components: project management success and project product success. Project management success primarily focuses on the processes applied to the project whereas project product success focuses on the overall effect of the project’s final product or output. There is a clear relationship between the two as both of them are inseparably linked and it would be foolish to believe that project management success does not influence the accomplishment of project product success.

The results within this article imply that ICT project performance rates have not improved over the past decade as they continue to fluctuate. ICT project success criteria have effectively remained the same from 2011 to 2013. However, the criterion of met business objectives rose significantly from third to last to third overall, signifying a paradigm shift in the way project success is perceived or measured. With regard to project success factors, communication between team and customers continues to emerge in the top three over the past decade. This affirms that effective communication is pivotal to project success.

This research postulated that by understanding the significance of each factor, it should be possible to predict whether an ICT project will succeed or fail. SEM was subsequently applied to construct a predictive model for ICT project success. The final model revealed that stakeholder management, communication and methodology are the driving factors behind ICT project success and that they are positively correlated to one another. Furthermore, the model empirically validated that ICT project success is dependent on project management success and project product success.

It was interesting that this research merely reiterated what has been known for decades, especially with regard to communication. The question should therefore be asked if any progress is being made to address the problem of poor ICT project performance rates and especially project success rates. Research within not only the ICT project management domain, but the greater project management domain could be seen to be stagnating and a fresh outlook is required if project management is to become a robust tool for creating economic and social wellbeing. Furthermore, maybe it is time to perform in-depth research into how ancient civilizations such as the ancient Egyptians successfully constructed the pyramids and the Sphinx. Learning from the past could create the stepping stone required to address the stagnating, uninspiring state of project management.

References

Ahlemann, F., Teuteberg, F., & Vogelsang, K. (2009). Project management standards—Diffusion and application in Germany and Switzerland. International Journal of Project Management, 27(3), 292-303.

Anglim, J. (2007). Structural equation modelling (Vol 2013).

Baccarini, D. (1999). The logical framework method for defining project success. Project Management Journal, 30(4), 25-32.

Bannerman, P. L. (2008). Defining project success: A multilevel framework. Paper presented at the Proceedings of the Project Management Institute Research Conference.

Belassi, W., & Tukel, O. I. (1996). A new framework for determining critical success/failure factors in projects. International Journal of Project Management, 14(3), 141-151. doi: http://dx.doi.org/10.1016/0263-7863(95)00064-X

Berlin, S., Raz, T., Glezer, C., & Zviran, M. (2009). Comparison of estimation methods of cost and duration in IT projects. Information and Software Technology, 51(4), 738-748. doi: http://dx.doi.org/10.1016/j.infsof.2008.09.007

Biggam, J. (2011). Succeeding with your master’s dissertation: A step-by-step handbook. New York: McGraw-Hill Education.

Boetticher, G. D. (2001). Using machine learning to predict project effort: empirical case studies in data-starved domains. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.19.111&rep=rep1&type=pdf

Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York: Guilford Press.

Bryde, D. (2008). Perceptions of the impact of project sponsorship practices on project success. International Journal of Project Management, 26(8), 800-809. doi: 10.1016/j.ijproman.2007.12.001

Camilleri, E. (2011). Project success: Critical factors and behaviours: Surrey, U.K.: Gower Publishing.

Cheng, M.-Y., & Wu, Y.-W. (2008). Dynamic prediction of project success using evolutionary support vector machine inference model. Paper presented at the Proceedings of the 25th International Symposium on Automation and Robotics in Construction.

Christensen, L. B., Johnson, R. B., & Turner, L. A. (2010). Research methods, design and analysis. Boston: Allyn & Bacon, Inc..

Cohen, D. J., & Graham, R. J. (2001). The project manager’s MBA: How to translate project decisions into business success. Hoboken, NJ: Wiley.

Cooke-Davies, T. (2002). The “real” success factors on projects. International Journal of Project Management, 20(3), 185-190. doi: http://dx.doi.org/10.1016/S0263-7863(01)00067-9

Crawford, L., & Pollack, J. (2007). How generic are project management knowledge and practice? Project Management Journal, 38(1), 87-96.

Crawford, L., Pollack, J., & England, D. (2006). Uncovering the trends in project management: Journal emphases over the last 10 years. International Journal of Project Management, 24(2), 175:184.

de Barcelos Tronto, I. F., da Silva, J. D. S., & Sant’Anna, N. (2008). An investigation of artificial neural networks based prediction systems in software project management. Journal of Systems and Software, 81(3), 356-367. doi: http://dx.doi.org/10.1016/j.jss.2007.05.011

de Wit, A. (1988). Measurement of project success. International Journal of Project Management, 6(3), 164-170. doi: http://dx.doi.org/10.1016/0263-7863(88)90043-9

Denscombe, M. (2010). The good research guide: For small-scale social research projects. New York: McGraw-Hill Education.

Doloi, H., Iyer, K. C., & Sawhney, A. (2011). Structural equation model for assessing impacts of contractor’s performance on project success. International Journal of Project Management, 29(6), 687-695. doi: http://dx.doi.org/10.1016/j.ijproman.2010.05.007

Downing, D., & Clark, J. (2010). Business statistics. Hauppauge, NY: Barrons Educational Series Incorporated.

Erasmus, W., & Marnewick, C. (2012). Project management—The saviour in turbulent times? A cross-sector analysis. Paper presented at the Proceedings of the 24th Annual Conference of the Southern Africa Institute for Management Scientists, Stellenbosch, South Africa.

Eveleens, J. L., & Verhoef, C. (2010a). The rise and fall of the Chaos report figures. IEEE Software, 27(1), 30-36.

Field, A. P. (2009). Discovering statistics using SPSS: (And sex and drugs and rock ‘n’ roll). Thousand Oaks, CA: Sage Publications.

Finnie, G. R., Wittig, G. E., & Desharnais, J. M. (1997). A comparison of software effort estimation techniques: Using function points with neural networks, case-based reasoning and regression models. Journal of Systems and Software, 39(3), 281-289. doi: http://dx.doi.org/10.1016/S0164-1212(97)00055-1

Foster, J. J., Barkus, E., & Yavorsky, C. (2006). Understanding and using advanced statistics: Thousand Oaks, CA: Sage Publications.

Fox, W., & Bayat, M. S. (2008). A guide to managing research. Claremont, South Africa: JUTA & Company Limited.

Gartner. (2013a). Gartner says cloud and CRM will drive enterprise software spending in 2013 and 2014. Retrieved 20 May, 2013, from http://www.gartner.com/newsroom/id/2358015

Gartner. (2013b). Forecast alert: IT spending, worldwide, 4Q12 Update Retrieved 25 March, 2013, from http://www.gartner.com/DisplayDocument?id=2291618

Gaskin, J. (2013a). SEM Series Part 3: Exploratory factor analysis (Vol. 2013).

Gaskin, J. (2013b). Exploratory factor analysis, Gaskination’s StatWiki Retrieved 5 August, 2013, from http://statwiki.kolobkreations.com/wiki/Exploratory_Factor_Analysis#Factor_Structure

Gaskin, J. (2013c). How to improve reliability (Cronbach’s Alpha) (Vol. 2013).

Gaskin, J. (2013d). SEM Series Part 5a: Confirmatory factor analysis (Vol. 2013).

Glesne, C., & Peshkin, A. (1992). Becoming qualitative researchers: An introduction. London: Longman.

Graham, J. M., Guthrie, A. C., & Thompson, B. (2003). Consequences of not interpreting structure coefficients in published CFA research: A reminder. Structural Equation Modeling, 10(1), 142-153.

Gray, A. R., & MacDonell, S. G. (1997). A comparison of model building techniques to develop predictive equations for software metrics. Information and Software Technology, 39, 425–437.

Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate Data Analysis (6th ed.). Upper Saddle River, NJ: Prentice-Hall International.

Hutcheson, G. D., & Sofroniou, N. (1999). The Multivariate Social Scientist: Introductory Statistics Using Generalized Linear Models: Thousand Oaks, CA: SAGE Publications.

Hyväri, I. (2006). Success of projects in different organizational conditions. Project Management Journal, 37(4), 31-41.

IDC. (2012). IDC Predicts Double-Digit Growth of IT Spending by South African Government Retrieved 21 February, 2013, from http://www.idc.com/getdoc.jsp?containerId=prZA23758612

Ika, L. A. (2009). Project success as a topic in project management journals. Project Management Journal, 40(4), 6-19. doi: 10.1002/pmj.20137

International Organization for Standardization. (2012). ISO 21500:2012 Guidance on project management. Geneva: International Organization for Standardization.

Jöreskog, K. G. (1993). Testing structural equation models. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 294-361). Newbury Park, CA: Sage Publications.

Jugdev, K., Perkins, D., Fortune, J., & White, D. (2013). An exploratory study of project success with tools, software and methods. International Journal of Managing Projects in Business, 6(3), 6-6.

Kaiser, H. (1974). An index of factorial simplicity. Psychometrika, 39(1), 31-36. doi: 10.1007/bf02291575

Karunanithi, N., Whitley, D., & Malaiya, Y. K. (1992). Using neural networks in reliability prediction, IEEE Software, 9(4), 53-59. doi: 10.1109/52.143107

Kendall, G. I., & Rollins, S. C. (2003). Advanced project portfolio management and the PMO: Multiplying ROI at warp speed. Plantation, FL: J. Ross Publishing, Incorporated.

Kim, D. Y., Han, S. H., Kim, H., & Park, H. (2009). Structuring the prediction model of project performance for international construction projects: A comparative analysis. Expert Systems with Applications, 36(2, Part 1), 1961-1971. doi: http://dx.doi.org/10.1016/j.eswa.2007.12.048

King, G., Keohane, R. O., & Verba, S. (1994). Designing Social Inquiry: Scientific Inference in Qualitative Research: Princeton, NJ: Princeton University Press.

Labuschagne, L., & Marnewick, C. (2009). The Prosperus report 2008: IT project management maturity versus project success in South Africa. Johannesburg, South Africa: Project Management South Africa.

Lee, S.-K., & Yu, J.-H. (2012). Success model of project management information system in construction. Automation in Construction, 25, 82-93. doi: 10.1016/j.autcon.2012.04.015

Lee, T. H., & Jung, S. (2000). Forecasting creditworthiness: Logistic vs. artificial neural net. The Journal of Business Forecasting Methods and Systems, 18(4), 28-30.

Lim, C. S., & Mohamed, M. Z. (1999). Criteria of project success: An exploratory re-examination, International Journal of Project Management, 17(4), 243-248. doi: http://dx.doi.org/10.1016/S0263-7863(98)00040-4

Ling, F. Y. Y., & Liu, M. (2004). Using neural network to predict performance of design-build projects in Singapore. Building and Environment, 39(10), 1263-1274. doi: http://dx.doi.org/10.1016/j.buildenv.2004.02.008

Markus, M. L., Axline, S., Petrie, D., & Tanis, C. (2000). Learning from adopters’ experiences with ERP: Problems encountered and success achieved, Journal of Information Technology, 15(4), 245-265.

Marnewick, C. (2012). A longitudinal analysis of ICT project success. Paper presented at the Proceedings of the South African Institute for Computer Scientists and Information Technologists Conference, Pretoria, South Africa.

Marnewick, C. (Ed.). (2013). Prosperus report—The African Edition. Johannesburg, South Africa: Project Management South Africa.

Marnewick, C., & Labuschagne, L. (2012). Factors that influence the outcome of ICT projects. Paper presented at the Proceedings of the Project Management South Africa Conference 2012, Johannesburg, South Africa.

Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor analysis to the study of self-concept: First- and higher-order factor models and their invariance across groups. Psychological Bulletin, 97(3), 562-582. doi: 10.1037/0033-2909.97.3.562

McKinney, V., Yoon, K., & Zahedi, F. M. (2002). The measurement of web-customer satisfaction: An expectation and disconfirmation approach. Information Systems Research, 13(3), 296-315. doi: 10.1287/isre.13.3.296.76

OECD. (2012). IT sector booms during downturn, says OECD. Retrieved 20 May, 2013, from http://www.oecd.org/internet/itsectorboomsduringdownturnsaysoecd.htm

Ohara, S. (2005). A guidebook of project & program management for enterprise innovation. Tokyo, Japan: Project Management Association of Japan.

Phillips, J. J., Bothell, T. W., & Snead, G. L. (2002). The project management scorecard. New York: Taylor & Francis.

Pinkerton, W. (2003). Project management: Achieving project bottom-line success. New York: McGraw-Hill Education.

Pinto, J. K., & Slevin, D. P. (1988). Critical success factors across the project life cycle. Project Management Journal, 19(3), 67-74.

Project Management Institute. (2013). A Guide to the Project Management Body of Knowledge: (PMBOK® Guide) – Fifth Edition. Project Management Institute, Incorporated.

Razi, M. A., & Athappilly, K. (2005). A comparative predictive analysis of neural networks (NNs), nonlinear regression and classification and regression tree (CART) models. Expert Systems with Applications, 29(1), 65-74. doi: http://dx.doi.org/10.1016/j.eswa.2005.01.006

Ricknas, M. (2013). IT investments will increase in 2013 thanks to software and device spending. Retrieved 20 May, 2013, from http://www.cio.com/article/725194/IT_Investments_Will_Increase_in_2013_Thanks_to_Software_and_Device_Spending

Roh, T. H., Ahn, C. K., & Han, I. (2005). The priority factor model for customer relationship management system success. Expert Systems with Applications, 28(4), 641-654. doi: http://dx.doi.org/10.1016/j.eswa.2004.12.021

Sekaran, U. (1992). Research methods for business: A skill-building approach: Hoboken, NJ: Wiley.

Sewchurran, K., Smith, D., & Roode, D. (2010). Toward a regional ontology for information systems project management. [Research Paper]. International Journal of Managing Projects in Business, 3(4), 681-692. doi: 10.1108/17538371011076118

Shang, J. S., Lin, Y. E., & Goetz, A. M. (2000). Diagnosis of MRSA with neural networks and logistic regression approach. Health Care Management Science, 3(4), 287-297.

Sheffield, J., & Lemétayer, J. (2013). Factors associated with the software development agility of successful projects. International Journal of Project Management, 31(3), 459-472. doi: 10.1016/j.ijproman.2012.09.011

Shenhar, A. J., Levy, O., & Dvir, D. (1997). Mapping the dimensions of project success. Project Management Journal, 28(2), 5-13.

Slevin, D. P., & Pinto, J. K. (1986). The project implementation profile: New tool for project managers. Project Management Journal, 17(4), 57-70.

Smith, D. C., Bruyns, M., & Evans, S. (2011). A project manager’s optimism and stress management and IT project success. [DOI: 10.1108/17538371111096863]. International Journal of Managing Projects in Business, 4(1), 10-27.

Sohn, S. Y., Kim, H. S., & Moon, T. H. (2007). Predicting the financial performance index of technology fund for SME using structural equation model. Expert Systems with Applications, 32(3), 890-898. doi: http://dx.doi.org/10.1016/j.eswa.2006.01.036

Sohn, S. Y., & Moon, T. H. (2003). Structural equation model for predicting technology commercialization success index (TCSI). Technological Forecasting and Social Change, 70(9), 885-899. doi: http://dx.doi.org/10.1016/S0040-1625(03)00004-0

Sonnekus, R., & Labuschagne, L. (2003). The Prosperus Report 2003. Johannesburg, South Africa: RAU Standard Bank Academy for Information Technology.

Stahl, D. (2008). Introduction to path analysis and structural equation modelling with AMOS (Vol. 2013).

Standish Group. (1995). Chaos Report. Boston: Author

Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics. New York: HarperCollins College Publishers.

Taylor, B., Sinha, G., & Ghoshal, T. (2006). Research methodology: A guide for researchers in management and social sciences. Delhi, India: PHI Learning.

Thomas, R. M. (2003). Blending qualitative and quantitative research methods in theses and dissertations. Thousand Oaks, CA: SAGE Publications.

Ullman, J. B. (1996). Structural equation modeling. In B. G. Tabachnick & L. S. Fidell (Eds.), Using multivariate statistics. New York: Harper Collins.

Vadapalli, A., & Mone, M. A. (2000). Information technology project outcomes: User participation structures and the impact of organization behavior and human resource management issues. Journal of Engineering and Technology Management, 17(2), 127-151. doi: http://dx.doi.org/10.1016/S0923-4748(00)00018-7

Van Der Westhuizen, D., & Fitzgerald, E. P. (2005). Defining and measuring project success. Paper presented at the European Conference on IS Management, Leadership and Governance 2005, Reading, United Kingdom. http://eprints.usq.edu.au/346/

Walliman, N. (2005). Your research project: A step-by-step guide for the first-time researcher. Thousand Oaks, CA: SAGE Publications.Wang, Y.-R., & Yu, C.-Y. (2011). Predicting project success using ANN-ensemble classificaiton models. Paper presented at the Proceedings of the 2011 IEEE 3rd International Conference on Communication Software and Networks (ICCSN)

Wateridge, J. (1998). How can IS/IT projects be measured for success? International Journal of Project Management, 16(1), 59-63. doi: http://dx.doi.org/10.1016/S0263-7863(97)00022-7

West, P. M., Brockett, P. L., & Golden, L. L. (1997). A comparative analysis of neural networks and statistical methods for predicting consumer choice. Marketing Science, 16(4), 370-391. doi: 10.1287/mksc.16.4.370

Westerveld, E. (2003). The project excellence model®: Linking success criteria and critical success factors. International Journal of Project Management, 21(6), 411-418.

Yatim, B. (2008). Statistical modelling (Special Topic: SEM) (Vol. 2013).

Yeo, K. T. (2002). Critical failure factors in information system projects. International Journal of Project Management, 20(3), 241-246. doi: http://dx.doi.org/10.1016/S0263-7863(01)00075-8

Young, R., & Poon, S. (2013). Top management support—Almost always necessary and sometimes sufficient for success: Findings from a fuzzy set analysis. International Journal of Project Management, 31(7). doi: http://dx.doi.org/10.1016/j.ijproman.2012.11.013

Zhang, G. P., Keil, M., Rai, A., & Mann, J. (2003). Predicting information technology project escalation: A neural network approach. European Journal of Operational Research, 146(1), 115-129. doi: http://dx.doi.org/10.1016/S0377-2217(02)00294-1

Nazeer Joseph holds a masters in commerce degree in information technology management and is a lecturer at the University of Johannesburg in South Africa. He lectures at the undergraduate level within the information technology domain. His research focuses on ICT project management with a particular focus on ICT project success.

Carl Marnewick, PhD, is head of the applied information systems department at the University of Johannesburg. He is a top-rated researcher and has presented several papers at local and international conferences. The focus of his research is the overarching topic and special interest of the strategic alignment of projects to the vision of the organizations. This alignment is from the initiation of a project to the realization of benefits. He is actively involved in the development of new international project management standards ISO 21500 (project management) and ISO 21502 (portfolio management).

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI or any listed author.

©2014 Project Management Institute Research and Education Conference

Advertisement

Advertisement

Related Content

Advertisement