### Project Complexityand Success in Complex Construction Projects
### Project Complexityand Success in Complex Construction Projects
Abstract: Although widely recognized both in literature and among practitioners, project complexity may cause poor project success, with
little empirical evidence supporting this contention. Therefore, this study analyzed, for the first time, the relationship between project
Downloaded from ascelibrary.org by Zhejiang University on 08/28/24. Copyright ASCE. For personal use only; all rights reserved.
complexity and success in complex construction projects and investigated how project complexity affects project success. First, project
complexity is hypothesized to be negatively related to project success. Second, on the basis of literature review and expert interviews,
a total of 245 questionnaire surveys on project complexity and project outcomes were collected in China. Project complexity was measured
as information, task, technological, organizational, environmental, and goal complexities by correlation and factor analyses. Finally, the
structural-equation modeling technique was used to test the hypothesis and explore the effect of different complexities on project success.
The findings of this study support the hypothesized negative relationship between the complexity and success of complex construction
projects. Furthermore, information complexity and goal complexity have significant negative effects on project success. The research would
have significant theoretical and practical significance for improving the theory of complex project management and achieving project
success in complex construction projects for project managers. DOI: 10.1061/(ASCE)ME.1943-5479.0000471. © 2016 American Society
of Civil Engineers.
Author keywords: Complex construction projects; Project complexity; Project success; Relationships; China.
Introduction components that may have the greatest potential effect on project
success.
The rapid rate of urbanization in recent years has resulted in an Studies have shown that project complexity affects project
increase in the number of complex construction projects in China, performance. However, in-depth analyses are lacking on specific
with large amounts of dollars invested in infrastructure construction relations and effects, thus resulting in poor application in prac-
(He et al. 2015; Hu et al. 2012). The lack of relevant knowledge on tice. First, existing papers on the relationship between project
the part of project managers often results in these projects being success and complexity mostly focus on the general project
beset with issues such as low performance, cost overruns, and complexity from an enterprise perspective and do not consider
schedule delays (Kennedy et al. 2011; Thomas and Mengel 2008). the characteristics of complex construction projects. There is a
Success is the ultimate goal of project management (Chan et al. particular lack of empirical study in China (He et al. 2015).
2004) and its factors play an important role in the planning, design, Second, the indicators of project complexity and success are
and construction of successful building projects (Parfitt and macroscopic and abstract; thus, applying the research results
Sanvido 1993). Not all complexity factors have a significant effect in practice is difficult. For example, in addition to the traditional
on project success. Thus, project managers should understand how golden triangle (quality, time, and cost), project success indica-
various project complexity components affect project performance. tors should also consider other successful statistics to provide
This approach can help project managers focus on the complexity references for project managers in handling projects (Macheridis
and Nilsson 2004).
1
Assistant Professor, School of Civil Engineering and Architecture, Accordingly, a total of 245 questionnaires about complex con-
Nanchang Univ., 999 Xuefu Ave., Honggutan New District, Nanchang struction projects were collected in China to empirically investigate
330031, P.R. China. the relationship between project complexity and success. The data
2
Professor, Research Institute of Complex Engineering and
were used to address the following questions:
Management, School of Economics and Management, Tongji Univ., No.
1, Zhangwu Rd., Shanghai 200092, P.R. China. • What are the key components of project complexity for complex
3
Ph.D. Student, Research Institute of Complex Engineering and construction projects?
Management, School of Economics and Management, Tongji Univ., No. • How does project complexity affect project success?
1, Zhangwu Rd., Shanghai 200092, P.R. China (corresponding author). • Do the various components of project complexity affect project
E-mail: [email protected] success equally? If not, how do they differ?
4
Ph.D. Student, Research Institute of Complex Engineering and The structural-equation model (SEM) is the main tool used
Management, School of Economics and Management, Tongji Univ., No. for path analysis and is unrestricted by regression analysis as-
1, Zhangwu Rd., Shanghai 200092, P.R. China. sumptions. SEM can not only handle multiple dependent vari-
5
Associate Professor, School of Tourism and Urban Management, ables at the same time but can also estimate factor structure and
Jiangxi Univ. of Finance and Economics, No. 169, East Shuanggang
factor relations (AMOS version 3.6). Project complexity and
Rd., Changbei, Nanchang 330013, P.R. China.
Note. This manuscript was submitted on December 30, 2015; approved success are composed of multiple dimensions and present a cer-
on May 3, 2016; published online on July 26, 2016. Discussion period open tain structure and gradation between each dimension. Therefore,
until December 26, 2016; separate discussions must be submitted for in- SEM is appropriate to use in this study to investigate the effect
dividual papers. This paper is part of the Journal of Management in En- of project complexity on project success in complex construc-
gineering, © ASCE, ISSN 0742-597X. tion projects.
2011; Gransberg et al. 2013; He et al. 2015). Baccarini (1996) clas- ity, health and safety, environmental performance, participants’ sat-
sified project complexity into organization complexity and technol- isfaction, user satisfaction, and commercial value. Yu et al. (2006)
ogy complexity. Tatikonda and Rosenthal (2000) believed that yielded five major categories, including project-related factors, hu-
project complexity is closely related to the interactions among man-related factors, process-related factors, input-related factors,
organizational elements and subtasks. Maylor et al. (2008) identi- and output-related factors. In China, Lin et al. (2005) divided the
fied the elements of project complexity as mission, organization, project success criteria into preliminary, construction, and opera-
delivery, stakeholders, and team. Brockmann and Girmscheid tion success criteria from the project lifecycle and all stakeholders’
(2008) divided complexity into five categories: task, society, cul- perspectives. Wang and Xu (2009) subdivided the indicators of
ture, operation, and cognition complexity. success into project implementation and project results. Project
As shown in the literature review of project complexity, each implementation mainly focuses on the objective control, interest
scholar has a different classification and perspective on project demands of participators, satisfaction, motivation, and harmony
complexity (He et al. 2012; Hu et al. 2012; Maylor et al. 2008; realization and is divided into 24 evaluation standards to measure
Remington and Pollack 2007; Tatikonda and Rosenthal 2000; Vidal the success of project delivery. Project results consider the exter-
and Marle 2008; Vidal et al. 2011). In the authors’ prior study as nality factors of fairness and sustainable development and the sat-
He et al. (2015), a six-category framework of project complexity isfaction of the whole process.
consisting of technological, organizational, goal, environmental, In conclusion, project success is very broad. It not only contains
cultural, and information complexities was proposed with a com- the project process but also includes the effect after project com-
prehensive literature review to measure the complexity of mega pletion. Project success can involve various project participants and
construction projects in China (He et al. 2015). However, the effect other stakeholders related to the project. A unified definition of
of the aforementioned complexity factors on project complexity is project success is not needed; thus, project success in this study
not fully understood and is still under investigation. Therefore, the refers to the success of the whole construction project. According
first objective of this study was to develop a measure for project to comprehensive existing research, the research of Chan et al.
complexity by examining the relationship between complexity fac- (2004) was chosen as a reference to the measures of project success
tors and project complexity. because of its comprehensive literature review and because it has
the most citations about success factors of construction projects.
They stated that the standard evaluation system for construction
Project Success project success includes time, cost, quality, health and safety, envi-
The concept of project success has remained ambiguously defined ronmental performance, participants’ satisfaction, user satisfaction,
in the construction industry (Chan et al. 2004; Joslin and Müller and commercial value.
2015). Stakeholders cannot achieve a consensus on project success
when it comes to project practices (Joslin and Müller 2015; Lim
and Mohamed 1999). Project success and project performance Hypothesis Development
are two similar but slightly different concepts, with project success A number of studies have tried to investigate the influence of
having a greater weight than project performance (Chan et al. project complexity on project performance. Puddicombe (2011)
2004). Project performance mainly focuses on the indicators of demonstrated via an analysis of more than 1,300 projects that tech-
project execution, but the time range of project success, such as nical complexity and novelty are important characteristics of a
the project’s influence after completion, is also included (Lim and project that have distinct effects on project performance. Antonia-
Mohamed 1999). dis et al. (2011) established in five case studies that the effects of the
The earliest criterion of project success is the golden triangle, socio-organo complexity of interconnections have similarities with
which consists of time, cost, and quality (Atkinson 1999; Jugdev the behavior of underdamped control systems and found that socio-
and Müller 2005; Molenaar et al. 2013). These measures are im- organo complexity is caused by interconnections that lead to a re-
portant but do not include the necessary measures of modern fac- duction in performance if not managed. Furthermore, Lebcir and
tors for project success (Atkinson 1999; Ika 2009). In addition to Choudrie (2011) built a project complexity framework for con-
the redefinition of project success, the roles and responsibilities of a struction projects and evaluated the effect of this framework on
project manager go far beyond the traditional golden triangle, in- project cycle time by using a system dynamics simulation model
cluding relations, cultural, and stakeholder management (Lam et al. that integrates project complexity, project operations, and time per-
2011; Meng et al. 2011; Ozorhon and Cinar 2015; Wong and formance. Tam (2010) assessed the effects of project technical com-
Cheung 2005). Bosch-Rekveldt (2011) stated that project success plexity on building production influences by using a clustering and
should be measured from internal and external perspectives. In ad- knowledge-based system. Bosch-Rekveldt (2011) concluded that
dition to guaranteeing project completion within a specified budget project complexity negatively influenced project performance in
Information complexity Information uncertainty (PC34); level of processing information (PC35); capacity of transferring information (PC36); degree
of obtaining information (PC37); integration of more than one system or platform (PC38); dependence of information system
(PC39); variety of language involved (PC40); number of countries or nationality involved (PC41)
Project success Time (PS1); cost (PS2); quality (PS3); health and safety (PS4); environmental performance (PS5); participants’ satisfaction
(PS6); user satisfaction (PS7); commercial value (PS8)
Software SPSS version 17.0 and AMOS version 16.0 were Table 3. Respondents’ Demographics
adopted for the analysis in this study; thus, a considerable number Characteristic Category Frequency Percentage (%)
of data had to be collected. However, scholars have not yet reached
a consensus for the sample number. For instance, Hou et al. Gender Male 188 76.7
Female 57 23.3
(2004) concluded that most of the SEM models need at least
Education Ph.D. 9 3.7
100–200 samples. Marsh et al. (1998), from the aspects of con- Master’s degree 60 24.5
vergence and fitting index, indicated that a greater number of sam- Bachelor’s degree 134 54.7
ples leads to better results. Hair et al. (1998) argued that sample Others 42 17.1
size is better than 200 but cannot become excessive because data Work experience ≤5 years 109 44.5
with more than 400 samples will result in a poor fitness index. Wu 6–10 years 77 31.4
(2010) believed that the best sample is 5 times the number of scale 11–20 years 47 19.2
questions. In this study, a total of 314 questionnaires were handed >20 years 12 4.9
out and 256 questionnaires were collected; thus, the recovery rate Designation Project manager 41 16.7
was 81.5%. After excluding 11 invalid questionnaires, 245 valid Department manager 25 10.2
Professional manager 55 22.4
questionnaires
Project engineer 73 29.8
were received; hence, the effective recovery rate was 78%. In Others 51 20.8
this sample, the majority of the respondents were male (76.7%) Project type Residential project 122 49.8
and possess a bachelor’s degree (54.7%). Details are given in Public project 81 33.1
Table 3. Industrial project 15 6.1
Others 27 11.0
Project size ≤10 million RMB 14 5.7
Operationalization of Variables 10–50 million RMB 34 13.9
50–100 million RMB 35 14.3
Project Complexity 100–500 million RMB 98 40.0
The operationalization of project complexity of information, task, >500 million RMB 64 26.1
technological, organizational, environmental, and goal complexity Project duration ≤12 months 20 8.2
was based on the questionnaire. A total of 41 questions were origi- 13–24 months 79 32.2
nally developed, reviewed, and validated through a series of data 25–36 months 92 37.6
>36 months 54 22.1
analyses (Luo et al. 2015).
First, correlation analyses of each item and of the total project
complexity were conducted to identify the key project complexity
factors of complex construction projects. Pearson correlation was the total project complexity is greater than 0.4 and significant at
used to explore the relationship between the project complexity fac- the 0.01 level (two-tailed). Thus, the variables related to the project
tors and the overall complexity and to investigate whether there is a complexity of complex construction projects include PC2, PC4,
significant correlation between them. The correlation coefficient PC7, PC8, PC10, PC11, PC12, PC13, PC14, PC15, PC16,
analysis between project complexity factors and the total complex- PC18, PC19, PC21, PC22, PC24, PC25, PC26, PC27, PC28,
ity needs to reach a significant level. Moreover, the correlation co- PC31, PC33, PC34, PC35, PC36, PC37, and PC39.
efficient should be at least 0.4 (Wu 2010). The correlation analysis The exploratory factor analysis method was subsequently
result between potential factors and total project complexity is adopted to classify the 27 key factors of project complexity for
shown in Table 4. Table 4 shows the correlation coefficient among complex construction projects. A principal component method was
PC2, PC4, PC7, PC8, PC10, PC11, PC12, PC13, PC14, PC15, conducted to extract the common factor and the maximum variance
PC16, PC18, PC19, PC21, PC22, PC24, PC25, PC26, PC27, method was used for factor rotation. A factor is extracted when the
PC28, PC31, PC33, PC34, PC35, PC36, PC37, and PC39, and eigenvalue is greater than 1 and vice versa (Wu 2010). This step
0.290a
0.006
0.000
PC20
PC41
0.174 significant at the 0.05 level. Tables 5 and 6 present the results of the
factor analyses and show the six factors to be extracted from
a
a
27 items of project complexity, with the eigenvalues cumulatively
0.000
0.000
PC19
PC40
0.529
0.335
explaining 59.230% of the total variance, which is close to 60%.
Factor loading of items was above 0.400, which indicates that the 6
a
a
0.000
0.000
PC18
PC39
0.489
0.406
extracted common factors can effectively reflect the 27 variables.
The grouping of variables was similar to the original concepts of
project complexity; therefore, the names of project complexity con-
a
a
0.000
0.000
PC17
PC38
0.364
0.380
a
0.000
0.000
PC16
PC37
0.417
0.541
Downloaded from ascelibrary.org by Zhejiang University on 08/28/24. Copyright ASCE. For personal use only; all rights reserved.
a
0.000
0.000
PC15
PC36
0.556
a
0.000
0.000
PC14
PC35
0.566
0.649
ing for project complexity and shows that all Cronbach’s alpha
coefficient values are greater than 0.60, thus indicating good inter-
nal consistency. With the exception of task complexity, the CITC
a
a
0.000
0.000
PC13
PC34
0.558
0.630
values were greater than 0.300, thus indicating that each item is
highly consistent with the sum of other items. Alpha if item deleted
a
a
0.000
0.000
is less than Cronbach’s alpha, thus indicating that the scale of in-
PC12
PC33
0.582
0.459
a
0.000
0.000
PC11
PC32
0.522
0.320
task complexity is 0.660, which is greater than 0.60; thus, the internal
consistency reliability is good. The CITC value of TAC4 is 0.227,
which is less than 0.300, and alpha if item deleted is 0.726, which
a
a
0.000
0.000
PC10
PC31
0.525
0.500
is greater than 0.660. Therefore, the item TAC4 was deleted, with the
result shown in brackets in Table 7. After deleting TAC4, the factor
a
0.000
PC30
0.390
0.382
PC9
a
0.000
0.000
PC29
0.408
0.379
PC8
indicating that each item is highly consistent with the sum of other
items. Alpha if item deleted is less than 0.726, thus indicating that the
Table 4. Correlation Analysis between Potential Factors and Total Project Complexity
a
0.000
0.000
PC28
0.431
0.532
PC7
Project Success
According to comprehensive existing research, the research of
a
a
0.000
0.000
PC27
0.283
0.466
PC6
Chan et al. (2004) was chosen as the reference for the project suc-
cess questionnaire. They proposed the construction project success
evaluation standard system, including time, cost, quality, health and
b
a
0.013
0.000
PC26
0.446
0.158
PC5
a
0.000
0.000
PC25
0.467
PC4
and shows that the overall alpha coefficient value of project success
scale is 0.837, which is greater than 0.8. The CITC values are be-
a
a
0.000
0.000
PC24
0.260
0.458
PC3
tween 0.404 and 0.650 (greater than 0.300), thus indicating that each
item is highly consistent with the sums of other items. Alpha if item
deleted values, with the exception of item PS8, are slightly greater
a
a
0.000
0.000
PC23
0.440
0.365
PC2
than 0.837. If item PS8 is deleted, then the internal consistency alpha
of the other items will increase but will be close to the original internal
a
consistency alpha value. The other items are less than 0.837; thus, the
Significance (two-tailed) 0.000
0.455
PC1
Results
Pearson correlation
Pearson correlation
Item
Table 6. Rotating Component Matrix for Key Factors of Project is reached. Therefore, significant differences exist between the
Complexity covariance matrix of the measurement model and the covariance
Factor matrix of empirical data; thus, other measures should be used to
comprehensively evaluate the fitting degree of the model. From
Item 1 2 3 4 5 6
the absolute fitting metrics, the goodness-of-fit index (GFI) value
PC13 0.821 0.048 0.015 0.231 0.024 0.012 is 0.841, the adjusted GFI (AGFI) value is 0.798 (slightly lower
PC12 0.784 0.034 0.034 0.190 0.081 0.054 than 0.90 standards), and the root-mean square error of approxima-
PC36 0.743 0.114 0.064 −0.251 0.152 0.223 tion (RMSEA) value is 0.048, which is less than 0.08. From the
PC37 0.726 0.051 0.110 −0.296 0.182 0.165
relative fitting index, the normal fit index (NFI) value is 0.799,
PC14 0.725 0.211 0.077 0.118 0.020 -0.020
PC35 0.701 0.160 0.167 −0.220 0.317 0.190 the incremental fit index (IFI) value is 0.909, and the comparative
PC10 0.665 0.049 0.053 0.127 0.056 0.192 fit index (CFI) value is 0.906; thus, both IFI and CFI are greater
PC34 0.464 0.097 0.379 0.038 0.370 0.121 than 0.90, with the exception of NFI. Given that no mature scale is
PC18 0.460 −0.221 0.383 −0.067 0.248 0.221 available for the project complexity of complex construction proj-
PC16 0.087 0.738 0.062 0.141 0.039 −0.045 ects, the measurement scales in this study are developed from
PC22 −0.028 0.638 0.342 0.267 −0.114 0.286 existing theory and interview results. Accordingly, the GFI, AGFI,
PC15 0.177 0.580 0.163 0.406 0.080 0.084 and NFI values are also acceptable. Therefore, the fitting index of
PC39 0.257 0.510 −0.123 −0.271 0.268 0.158
project complexity and project success (PC&PS) model can be re-
PC26 −0.045 0.160 0.668 0.160 0.207 0.104
PC24 0.007 0.435 0.622 −0.041 −0.020 0.275 garded as good.
PC25 0.450 0.055 0.594 0.009 0.012 −0.134 A summary of the standardized coefficients of the final model is
PC19 0.367 −0.085 0.511 0.026 0.317 0.033 shown in Table 10. All path coefficients are positive and significant
PC7 0.050 0.171 0.084 0.803 0.184 0.095 at p < 0.05; thus, their significance to the model is augmented. In
PC8 0.020 0.185 0.045 0.745 0.135 0.218 the SEM model, project complexity was classified as information
PC27 0.169 −0.099 0.174 0.148 0.785 0.019 complexity, task complexity, technological complexity, organiza-
PC28 0.202 0.108 −0.008 0.063 0.675 0.322 tional complexity, environmental complexity, and goal complexity;
PC31 −0.013 0.376 0.279 0.217 0.457 0.085
project success was directly reflected by eight project targets,
PC33 0.107 0.267 0.343 0.193 0.453 −0.189
PC4 0.090 0.311 0.044 0.224 0.116 0.583 namely, time, cost, quality, health and safety, environmental perfor-
PC11 0.452 –0.147 0.028 0.250 0.106 0.536 mance, participants’ satisfaction, user satisfaction, and commercial
PC2 0.452 –0.113 0.174 –0.198 0.007 0.483 value.
PC21 0.208 0.245 0.103 0.224 0.137 0.472
Note: The principal component extraction method and the Kaiser
standardization rotation method were used, as well as orthogonal rotation
Relationships between Different Complexity and
after 14 times the iterative convergence. The bold text signifies that the Project Success
items were classified into different factors. Different complexities have various effects on project success.
SEM was then used to analyze the influence of different complex-
ities on project success, including information complexity, task
SEM model, whereas the key factors of project complexity are the complexity, technological complexity, organizational complexity,
dominant variables. The testing result of the theoretical model is environmental complexity, and goal complexity. The SEMs are
shown in Fig. 1 and Tables 9 and 10. The analysis result shows shown in Figs. 2–7.
that the path coefficient of project complexity to project success Table 9 presents the results of GOF measures and shows that the
is −0.254 and is significant at p < 0.001, thereby supporting hy- model fitting χ2 =df is less than 3, p ¼ 0.000, and a significant
pothesis H0. It is suggested that project complexity has a significant level is reached. Therefore, significant differences exist between
negative correlation with project success in Chinese complex con- the covariance matrix of the measurement model and the covari-
struction projects. ance matrix of empirical data. Therefore, other measures should be
The goodness-of-fit (GOF) indices are essential tools for assess- used to comprehensively evaluate the fitting degree of the model.
ing the fitness of SEMs. Table 9 shows that the model fitness of the From the absolute fitting metrics, the GFI and AGFI values are
final SEM for project complexity and project success is supported close to or greater than 0.90; the RMSEA value is 0.048, which is
by the results of the indices. The results of the model fitting show less than 0.08. From the relative fitting index, the NFI, IFI, and CFI
that χ2 =df is 1.667 (less than 3), p ¼ 0.000, and a significant level values are almost greater than 0.90. Accordingly, the results
.10 e5 IC5
.38 Information
e35
.78 .62 complexity
-.16 e6 IC6 .88
.12 .60
.36
.12 e7 IC7
.65
.42
e8 IC8 .55
.13
.30
e9 IC9
e36
-.48 -.12 .25 .17
e10 TAC1 .50
-.36 .54
-.18 -.20
e11 TAC2
.29 Task
.72
complexity
-.11
.19 .48 e12 TAC3 .85 .32
PS1 e27
Downloaded from ascelibrary.org by Zhejiang University on 08/28/24. Copyright ASCE. For personal use only; all rights reserved.
.90
.32 .16 .64 .49
.57 .32
.20 .27 e13 TEC1 .40 PS2 e28
.34 .59 .41
-.32 Technological .56
-.24.27
.23 e14 TEC2 e37 e41 .53
-.17 e15 TEC3
.16 complexity .73 PS3 e29
-.31 -.18 .40 .80 .06 .24
.27 .39 .63 .55 .30
.16 e16 TEC4 .07 Project -.25 Project PS4 e30
-.18 .14 -.88 .22
.15 .84
e17 OC1 .92 complexity success PS5 .37
e31 -.36 .35
.23 -.17
.56
Organizational .26 .61
.30 e38 .72
e18 OC2 .75
complexity .85 PS6 e32 .28
.42 .66 .66 .44
.21 .19 e19 EC1 .65 .43 PS7 e33
.37
.59
e20 EC2 .77 .85
Environmental .14
.21 PS8 e34
.16 -.53 e21 EC3 complexity
.46
.36 .60
.15 e22 EC4
e39
.15
e23 GC1 .38 .73
-.22
.38 .62
e24 GC2 Goal
.10 .35 e40
e25 GC3 complexity
.60
-.22 .21
GC4 .46
e26
Discussions consistent with Lebcir and Choudrie (2011), who simulated the
effects of project complexity on time to complete construction proj-
The measurement of project complexity developed in this study ects. Bosch-Rekveldt (2011) also confirmed that project complex-
was therefore thoroughly verified and checked for completeness, ity generally is assumed to decrease project performance, and
resulting in an updated version of the framework for complex con- Williamson (2011) showed that project complexity is negatively re-
struction projects. Project complexity was classified as information lated to IT project success. To address project complexity, the client
complexity, task complexity, technological complexity, organiza- could adopt the program management approach to simplify the
tional complexity, environmental complexity, and goal complexity. complexities and sustain the control of the dispensed execution
The measurement model of project complexity can be used to iden- of the project (Hu et al. 2014). Remington and Pollack (2007)
tify and measure the key factors, so as to give references for man- stated that program management is a pragmatic means of dealing
aging project complexity. The research investigated the relationship with nearly all types of project complexity.
between project complexity and project success, which is helpful Information complexity has a significant negative effect on
for achieving project success in complex construction projects for project success (standardized coefficient ¼ −0.312). Senescu et al.
project managers. (2014) has proved that project complexity affects collaboration,
The SEM results suggested that project complexity has a sig- sharing, and understanding; therefore, the industry will need to
nificant negative correlation with project success. The result is consider improving the timing and the approach taken to implement
TEC2 ← technological complexity 0.585 that technological complexity and novelty are important charac-
TEC1 ← technological complexity 0.396 teristics of a project that have distinct effects on project perfor-
OC2 ← organizational complexity 0.751 mance. Accordingly, project managers need to explicitly make
OC1 ← organizational complexity 0.915 an analysis of technical complexity and novelty and make them
EC4 ← environmental complexity 0.598 part of their planning calculus if superior performance is to be
EC3 ← environmental complexity 0.456 achieved (Puddicombe 2011).
EC2 ← environmental complexity 0.771 Organizational complexity also has no significant effect on
EC1 ← environmental complexity 0.647
project success (standardized coefficient ¼ 0.103). This research
GC4 ← goal complexity 0.459
GC3 ← goal complexity 0.595 result contrasts with that of Antoniadis et al. (2011), who found
GC2 ← goal complexity 0.618 that socio-organo complexity is caused by interconnections that,
GC1 ← goal complexity 0.385 if not managed, could lead to a reduction in performance. The
IC9 ← information complexity 0.546 reason is that organizational complexity in this study is primarily
IC8 ← information complexity 0.647 identified as organization structure, including the number of organi-
IC7 ← information complexity 0.602 zational structure hierarchies, organizational units, and depart-
IC6 ← information complexity 0.882 ments. However,the level of training of project team members and
IC5 ← information complexity 0.616
implementation of appropriate actions were selected by Antoniadis
IC4 ← information complexity 0.735
IC3 ← information complexity 0.786
et al. (2011) as the influence factors for project performance, which
IC2 ← information complexity 0.705 is consistent with the relationship between information complexity
IC1 ← information complexity 0.617 and project success in this study.
PS1 ← project success 0.570 An insignificant negative correlation was found between envi-
PS2 ← project success 0.563 ronmental complexity and project success (standardized coefficient ¼
PS3 ← project success 0.729 −0.110). Bosch-Rekveldt (2011) reached the same conclusion that
PS4 ← project success 0.552 the negative correlations between project complexity elements and
PS5 ← project success 0.608 project performance were found in the areas of interfaces between
PS6 ← project success 0.849
different disciplines and a lack of company internal support.
PS7 ← project success 0.663
PS8 ← project success 0.368
Significant negative correlations between goal complexity and
project success (standardized coefficient ¼ −0.231) were also
a
All standardized coefficient values are significant at p < 0.05. The bold found through the SEM result. The result was also proved by
text signifies the different dimensions of project complexity and the Bosch-Rekveldt (2011), who found the strongest correlations
relationship between project complexity and project success.
between project complexity elements and project performance
(negative) in the areas of goals and scope. According to goal com-
appropriate techniques (Antoniadis et al. 2011). According to the plexity, a client organization could apply the project breakdown
factors of information complexity identified in this study, the client structure and work breakdown structure (PBS/WBS) tools to align
could establish a separate communication management system to the tasks of different organizational units and the overall objectives
promote and integrate communication activities among designers, of complex construction projects; some divisions such as the cost
.37
e1 IC1
.58 .24
.37 e2 IC2 .60 PS1 e10
.49 .54
.61 e18 .25
.17 e3 IC3 .76 .50 PS2 e11 .22
.54 .78 .23 .29
.53 .10 .48 PS3 e12
-.72 e4 IC4 .73
.60 .36 .18
.36
.41 .64 Information -.31 Project PS4 e13 .15.08
.12 .21 e5 IC5 .43 .04 -.17 .26
.81
complexity success PS5 e14 -.35.22.37
e6 IC6 .90 .66 .63 .04 -.15
.34 .58
.79 PS6 e15 .14-.17
e7 IC7 .77 .60 -.16
.60 .34
PS7 e16 .22
.16e8 .36 .12
IC8 PS8 e17
.20 .26 .51
e9 IC9
.32
Downloaded from ascelibrary.org by Zhejiang University on 08/28/24. Copyright ASCE. For personal use only; all rights reserved.
PS1 e5
.56 .31 .50
.32
e13 .56
PS2 e6
.54
e1 TEC1 .57 .00 .74 PS3 e7
.28 .56 .31 .23
.53 PS4 e8
e2 TEC2 Technological -.01 Project .36 .14 -.96 .22
.37 complexity PS5 e9
success .60
-.39 .35
e3 TEC3 .73
.28 .61.53 .86 PS6 e10 .29
.43
e4 TEC4 .66 PS7 e11
.36 .13
PS8 e12
.32
PS1 e3
.56 .31 .50
e11 PS2 e4
.56 .55
.01 .74 PS3 e5
.49.70 .56 .31 .23
e1 OC1 Organizational .10 Project PS4 e6
1.00 .36 .14 -.97 .22
e2 OC2 complexity success PS5 e7 -.39 .36
1.00 .60
.74
.86 PS6 e8 .30
.65 .43
PS7 e9
.36 .13
PS8 e10
.32
PS1 e5
.38 .56 .31 .50
e13 .56 PS2 e6
e1 EC1 .55
.63 .01 .74 PS3 e7
.62 .31 .23
.80 .56
e2 EC2 Environmental -.11 Project PS4 e8
.21 .36 .14 -.98 .22
complexity success PS5 e9 -.39 .36
.60 .74
-.63 e3 EC3 .45
.40 .63 .86 PS6 e10 .30
.43
e4 EC4 .65 PS7 e11
.36 .13
PS8 e12
.32
PS1 e5
.57 .32 .49
.13 PS2 e6
e13 .56
.55
e1 GC1 .05 .74 PS3 e7
.66 .81.36 .55 .30 .24
e2 GC2 Goal -.23 Project PS4 e8
.21 .37 .14 -.93 .22
.23 complexity success PS5 e9 -.36 .35
e3 GC3 .48 .60
.34 .72
.11 .85 PS6 e10 .29
e4 GC4 .43
.66 PS7 e11
.36 .13
PS8 e12
on complex networks.” Sci. Technol. Progr. Policy, 25(11), 5–8 (in ships between project complexity and communication.” J. Manage.
Chinese). Eng., 10.1061/(ASCE)ME.1943-5479.0000121, 183–197.
Lim, C., and Mohamed, M. Z. (1999). “Criteria of project success: An SPSS version 17.0 [Computer software]. SPSS Inc., Prentice Hall
exploratory re-examination.” Int. J. Project Manage., 17(4), 243–248. International, New Delhi, India.
Lin, M., Shen, L., Ma, S. H., and Chen, J. H. (2005). “System thinking of Tam, C. (2010). “Assessment of impacts of project technical complexity on
project success criteria based on the life cycle.” J. Ind. Eng. Manage., building production using clustering and knowledge-based system.”
10(1), 101–105 (in Chinese). Proc., CIB W78 2010, Tsinghua University Press, Beijing.
Tatikonda, M. V., and Rosenthal, S. R. (2000). “Technology novelty,
Linstone, H. A., and Turoff, M. (1975). The Delphi method: Techniques
project complexity, and product development project execution success:
and applications, Addison-Wesley, Reading.
A deeper look at task uncertainty in product innovation.” IEEE Trans.
Liu, A. M. (1999). “A research model of project complexity and goal com-
Eng. Manage., 47(1), 74–87.
mitment effects on project outcome.” Eng. Constr. Archit. Manage.,
Thomas, J., and Mengel, T. (2008). “Preparing project managers to deal
6(2), 105–111.
with complexity: Advanced project management education.” Int. J.
Ludwig, B. (1997). “Predicting the future: Have you considered using the
Project Manage., 26(3), 304–315.
Delphi methodology.” J. Extension, 35(5), 1–4.
Vidal, L. A., and Marle, F. (2008). “Understanding project complexity:
Luo, L., He, Q. H., and Shu, L. L. (2015). “Identifying the project complex- Implications on project management.” Kybernetes, 37(8), 1094–1110.
ity factors of complex construction projects.” 2015 Int. Conf. on Man- Vidal, L. A., Marle, F., and Bocquet, J. C. (2010). “Using a Delphi process
agement Science and Engineering, IEEE Technology Management and the analytic hierarchy process (AHP) to evaluate the complexity of
Council, NJ. projects.” Expert Syst. Appl., 38(5), 5388–5405.
Macheridis, N., and Nilsson, C. H. (2004). “Managing project complexity: Vidal, L. A., Marle, F., and Bocquet, J. C. (2011). “Measuring project com-
A managerial view.” Lund Institute of Economic Research. plexity using the analytic hierarchy process.” Int. J. Project Manage.,
Marsh, H. W., Hau, K. T., Balla, J. R., and Grayson, D. (1998). “Is more 29(6), 718–727.
ever too much? The number of indicators per factor in confirmatory Wang, J., and Xu, Y. J. (2009). “Large engineering project success criteria
factor analysis.” Multivariate Behav. Res., 33(2), 181–220. under the vision of system thinking.” J. Chinese Eng. Sci., 10(12),
Maylor, H., Vidgen, R., and Carver, S. (2008). “Managerial complexity in 79–85 (in Chinese).
project based operations: A grounded model and its implications for Williamson, D. J. (2011). “A correlational study assessing the relationships
practice.” Project Manage. J., 39(S1), S15–S26. among information technology project complexity, project complica-
McComb, S. A., Green, S. G., and Compton, W. D. (2007). “Team flex- tion, and project success.” Capella Univ., Minneapolis.
ibility’s relationship to staffing and performance in complex projects: Wong, P., and Cheung, S. (2005). “Structural equation model on trust and
An empirical analysis.” J. Eng. Technol. Manage., 24(4), 293–313. partnering success.” J. Manage. Eng., 10.1061/(ASCE)0742-597X
Meng, X., Zhao, Q., and Shen, Q. (2011). “Critical success factors for (2005)21:2(70), 70–80.
transfer-operate-transfer urban water supply projects in China.” J. Man- Wu, M. L. (2010). The questionnaire analysis—SPSS statistical practice
age. Eng., 10.1061/(ASCE)ME.1943-5479.0000058, 243–251. and application, Chongqing Univ. Press, Chongqing (in Chinese).
Mihm, J., Loch, C., and Huchzermeier, A. (2003). “Problem-solving oscil- Xia, B., and Chan, A. P. C. (2012). “Measuring complexity for building
lations in complex engineering projects.” Manage. Sci., 49(6), 733–750. projects: A Delphi study.” Eng. Constr. Archit. Manage., 19(1), 7–24.
Molenaar, K. R., Javernick-Will, A., Bastias, A. G., Wardwell, M. A., and Yu, A. T. W., Shen, Q., Kelly, J., and Hunter, K. (2006). “Investigation of
Saller, K. (2013). “Construction project peer reviews as an early indi- critical success factors in construction project briefing by way of con-
cator of project success.” J. Manage. Eng., 10.1061/(ASCE)ME.1943 tent analysis.” J. Constr. Eng. Manage., 10.1061/(ASCE)0733-9364
-5479.0000149, 327–333. (2006)132:11(1178), 1178–1186.