Applicability of Integrated Project Delivery Principles Based on a Measurement Model in China

Abstract

This paper explores a novel approach to clarify the applicability of the Integrated Project Delivery (IPD) principles in a certain region or country which has a specific cultural background, the paper further promotes the effective application of IPD principles. The method of combining qualitative and quantitative analysis is adopted, incorporating scientometric analysis, questionnaire survey, qualitative analysis, and a measurement model. On the basis of 310 sample data collected from various practitioners and researchers in China’s construction industry, the measurement models of IPD applicability analysis on the average level of sample (ALS) and the best level of sample (BLS) are established, respectively. The results show that these measurement models are essentially consistent with the composition of the conceptual model, including project performance, contractual principles, collaboration-related principles, behavioral principles, and catalysts. Thus, this paper provides a scientific and methodological basis for how to effectively apply the IPD principles. This study sheds light on project delivery system research in two aspects: it proposes a new idea to study the applicability of IPD principles and lays a foundation for the future quantitative study on the impact mechanism of IPD principles on project performance.

1. Introduction

Referring to the traditional construction process, the stages of design, procurement, construction, and delivery are relatively separate, and there is a lack of collaboration among them. Thus, low productivity and high social cost are common problems [1]. As the development product of the traditional delivery mode tends towards a more integrated aspect, integrated project delivery (IPD) is well-known for its efficient collaboration, which eliminates the drawbacks of Design–Bid–Build and Design–Build [2]. IPD is considered as an effective method to overcome the fragmentation in traditional construction [3]. Moreover, the application of IPD in traditional contracting systems can overcome inefficiency issues and promote projects’ success [4]. The close relationship between IPD and high-performance buildings is detailed as follows: the realization of high-performance buildings depends on the integration of building systems; the realization of an integrated building system depends on the integration of various processes; the integration of each teams determines the realization of the integration process; the integration of various information affects the implementation of integration teams, that is, the efficient realization of Building Information Modeling (BIM) +; and simulation and visualization are the main ways for the integration team to communicate with BIM + information technology. Furthermore, collaboration and the co-location of the team are the main ways of ensuring team integration; the production management method enables the design, assembly, and construction of integrated building systems to be effectively realized, and the quantitative index defines the performance of the building and affects integration. The above content should be supported by the contract or framework [1]. Thus, one of the necessary conditions for IPD to achieve efficient collaboration is to apply BIM technology. However, one of the main reasons for the slow promotion of BIM technology is that the traditional project management method cannot maximize the benefits of BIM. Therefore, IPD is one of the best choices to reflect this collaborative idea. The coupling of BIM and IPD mutually facilitates the effective use of the each other [5].

The combination of IPD and the integrated practices needs further theoretical research [6]. IPD has various definitions, which usually are defined by some scholars as a delivery method [7,8] or a philosophy that can be applied to various types of delivery methods [9,10,11]. Due to the limitations in management, technology, and law, the application of pure IPD is restricted in most projects [12]. Add to that, IPD, deployed as a delivery method, has not yet found a wide application and is currently dominated by case studies of similar pilots or specific purposed cases [12,13]. Nonetheless, most researchers still insisted that a more flexible way to improve collaboration is to apply IPD as a philosophy [14]. Thus, IPD is characterized by the implementation of a series of basic principles [15]. When IPD is used as a philosophy rather than as a delivery method, some principles may be effective without formal contract languages and enforcements; however, when IPD plays the role of a delivery method [11], these assumptions no longer hold. IPD philosophy can be combined with traditional project delivery systems, which will contribute to building trust and cooperation and promoting integrated collaboration [16]. The level of collaboration is determined by the quantity and degree of application of IPD principles [14]. Even though IPD principles are classified differently, the core idea focuses on collaboration. The evolution of IPD principles has been shared in all American Institute of Architects (AIA) publications from 2006 to 2012 [10,17,18], in which a report from the National Association of State Facilities Administrators develops IPD principles into 15 more detailed principles, which are partitioned into contractual and behavioral principles, as well as catalysts or tools for IPD [11]. On the basis of these 15 IPD principles, the collaboration was divided into three levels, including typical collaboration (non-IPD), enhanced collaboration (IPD-ish), and required collaboration (IPD). Both levels one and two referred to the projects that apply IPD as a philosophy, while level three referred to projects that use IPD as a delivery method (pure IPD). Owners who are considering IPD delivery but are unable or unprepared to launch a “pure” IPD project may choose to examine IPD as a philosophy as a backup plan [19]. However, all these principles cannot be implemented in all project contexts at national or regional levels. Nevertheless, those who implement some but not all of the principles may still deliver much of IPD’s values. Therefore, it is very important to study the applicability of IPD principles in specific cultural backgrounds, regions, or countries where the conditions for using IPD as a delivery method are still not applied.

Given the high-development speed of construction, the increasing demand for informatization in building industry, and the obstacles for a direct introduction of pure IPD, China is currently one of the typical countries that can enhance the use of IPD as a philosophy. Considering China as an example, this paper explores a way to clarify the applicability of the IPD principles in a certain region or country, further providing a scientific reference for the efficient application in a certain region or country where IPD is deployed as a philosophy.

In terms of research methods, this paper adopts mainly the combination of qualitative and quantitative methods, including scientific metrological analysis, questionnaire survey, qualitative analysis, and a measurement model. In terms of research content, the conceptual model is proposed on the basis of the literature, and the measurement models of two groups of samples are established via the questionnaire method. The conceptual model is verified by the measurement model, and then the applicability of IPD principles in China is analyzed. The research process includes the foundation of the research (shown in Part 2), the research methodology (presented in Part 3), the results and analysis (refer to Part 4), the discussion (shown in Part 5) and, finally, the conclusion and some future works (provided in Part 6). In more detail, the foundation of the research is mainly: (1) proving the necessity of applying the IPD principles by scientometric analysis and summarizing the research results directly related to the application of IPD through an inductive literature method; and (2) determining whether a region or country with a specific cultural background is suitable for the application of IPD as a philosophy or a delivery method via the first questionnaire survey. Then, on the basis of the conceptual and the measurement model, the paper analyzed the results: (1) building the conceptual model for IPD principles by applying analysis through a qualitative analysis; and (2) establishing the two sets of measurement models concerning the applicability analysis of IPD on the basis of the second questionnaire survey and measurement model. The last section of this paper reveals the internal relationship of IPD principles in the application process via the comparative analysis of two sets of measurement models.

2. Foundation of the Research

2.1. Scientometric Analysis of the Related Literature about IPD Principles

Referring to the Web of Science database, 288 publications with the themes of “integrated project delivery principles” and “IPD principles” were obtained. After data screening, articles with repeated contents or having different meanings than the keywords were excluded, and only 14 articles related to the previously mentioned themes were maintained. The study conducted scientometric analysis by VOS viewer. The size of the circle in the map indicates the frequency of the keywords’ co-occurrence. The larger the circle is, the higher the frequency of keywords’ co-occurrence will be. The line represents the connection strength of the keyword. The thicker the line is, the stronger the relationship between the two keywords will be. In addition, the higher the connection strength is, the greater the impact on the co-occurrence network will be. IPD was the most frequently cited research keyword (as shown in Figure 1). Other common keywords in integrated project delivery involve IPD principle, implementation, application, model, barrier, collaboration, and BIM. Thus, it is necessary to analyze the applicability of IPD principles.

The research directly related to the literature on IPD principles is very limited. For instance, one can list the following: (1) IPD principles need to be applied to other technologies. Ahmed et al. [20] held that cooperative game theory and IPD principles can be applied together in construction bidding to achieve the goals of all parties. Through an IPD assessment model of public building projects, Ahmad et al. [14] considered that information and communication technology tools can foster IPD principles. The application of IPD contracts and the resolution of other cultural and organizational paradigms will facilitate gaining sway and establishing a presence in the construction sector; (2) IPD principles have an impact on project performance or results. Zhang and Hu [15] discussed how to apply IPD principle to achieve better results. Compared with non-IPD projects, the overall performance of IPD or near-IPD projects has a significant statistical significance [21]. In addition, Andary et al. [2] proved that the application of IPD principles has improved the performance of the case study project by 1.3 times. The integrated application of IPD, lean principles, and BIM can effectively improve schedule performance [22]. In traditional delivery methods, IPD-ish advocates the application of IPD principles with some limited risk-sharing ways, namely those methods using the IPD concepts and its philosophy without referring to a multiparty contract [9]. Projects based on IPD-ish are less effective at solving cost performance and change orders than dealing with scheduled performance and defects [23]. However, built on the properties of the project and the specific improvement of key performance indicators, it can be conducive to the high cost-effective implementation of some IPD principles. The implementation of IPD concepts in specific projects was, therefore, rarely addressed in earlier research [15]. In addition, there is virtually little study on how to apply IPD concepts to real building projects.

2.2. Verification That China Is Suitable for Adopting IPD as a Philosophy

According to the collaboration level, 15 IPD principles proposed by NASFA et al. are used in the paper [11]. Given the IPD categorization method presented by Boodai [18], as well as the feasibility of the above IPD principles, the paper assumes that China is currently suitable for IPD as a philosophy. In order to verify this hypothesis, this paper roughly investigates, as the first step, the current application of IPD principles and the BIM technology via field visits, then makes an in-depth investigation through the first questionnaire (interested readers can refer to it on the following website https://www.wjx.cn/vm/waUUwNB.aspx (accessed on 31 October 2022)). Then, some aspects, related to the first questionnaire, were identified: (1) type of company (universities or research institutes; construction units; design institutes; consulting units; owners; suppliers; and others); (2) years of experience (<3 years; 3~5 years; 5~8 years; and >8 years); (3) level of knowledge of IPD (experienced with IPD; inexperienced though informed about IPD; and inexperienced and unfamiliar with IPD); (4) number of BIM projects in which one participated (never; 1~2; 3~5; 6~10; and more than 10); and (5) type of collaboration level at which one thinks China currently is, i.e., typical collaboration, enhanced collaboration, and required collaboration (as shown in Figure 2). By examining different publication’s datasets and participating in relevant conferences and forums concerning the construction industry, we collected the email addresses of practitioners and experts in relevant fields, inviting them to respond to the questionnaire on the network platform. Using university resources as a guide, we sent the invitation to the personnel of the professional practice base and to recent graduates working in the specialization to complete the questionnaire. To sum up, 200 questionnaires were distributed, and 168 were received. Thus, 84% of the total data was retrieved.

The results of the various questionnaires show that the company type of the respondents are as follows: universities or research institutes (25%), construction units (22.62%), design institutes (22.62%), consulting units (17.86%), owners (8.33%), suppliers (0.60%), and others (2.97%). The results are in conformity with the conclusions of Mei et al. (2017), namely, the owners with the greatest benefit are less motivated to implement BIM and IPD than universities or research institutes, design institutes, and contractors. In addition, 51.19% of the respondents have participated in three or more BIM projects, of which 39.88% have six years or more working experience, indicating that experienced practitioners or researchers have more actively participated in BIM projects. Additionally, 36.90% of the respondents consider that they are inexperienced and unfamiliar with IPD, indicating that most practitioners and researchers actually know of, have heard of, or even had IPD project experience. The proportion of respondents who believe that China is ready for typical collaboration and enhanced collaboration is 88.09%; this shows that most practitioners or researchers believe that China is more suitable to take IPD as a philosophy now. Moreover, BIM technology is gradually popularized in China, and IPD principles have been applied in practical engineering projects. Although the application of some IPD principles in China’s construction industry is limited, researchers and practitioners still have a strong interest in IPD principles.

2.3. Establishment of the Conceptual Model for IPD Principle Application Analysis

Using IPD as a philosophy or a delivery method on any project requires a series of basic principles that need to be addressed. According to the classification from NASFA et al. [11], the effective application of contractual principles, behavioral principles, and catalysts can improve any IPD project. Contractual principles refer to those that may be written into agreements, aiming to improve integration and build a trusted working environment by the allocation of the project’s risks and benefits among all parties. The behavioral principles are those that are selection-based and necessary for project optimization. The behavioral principles, including the rules of integration and communication for all parties, are important factors affecting the efficiency and the performance of different teams. Finally, the catalysts are those that can be very useful for optimizing project results.

To highlight the advantages of BIM in communication, this paper considers the rate of requests for information as the main indicator to assess the communication cost. Generally, schedule, cost, and quality are the iron triangle indicators to measure the construction project performance. Thus, project performance is used to measure the application effect of IPD principles, including Construction Cost (CC1), Schedule Cost (SC), Quality Cost (QC), and Communication Cost (CC2). In addition, some references support the idea that IPD principles affect the project performance. Thus, this paper proposes a conceptual model of application analysis of IPD principles (please refer to Figure 3 and

Table 1. Summary of latent variables and observed variables.

Latent Variable/Category/Factor	Observed Variable	Item/Indicator	Reference
Contractual principles	x1	KPBTE: key participants bound together as equals	[11,17]
	x2	LWKP: liability waivers between key participants waivers
	x3	EIKP: early involvement of key participants
	x4	FT: fiscal transparency among key participants
	x5	JDPTC: jointly developed project target criteria
	x6	SRR: shared financial risk and reward based on project outcome
	x7	ID: intensified design
	x8	CDM: collaborative decision-making
Behavioral principles	x9	MRT: mutual respect and trust
	x10	WTC: willingness to collaborate
	x11	OC: open communication
Catalysts	x12	MA: multiparty agreement
	x13	BIM: building information modeling
	x14	LC: lean design and construction
	x15	CT: co-location of team
Project performance	x16	CC1: projects can be completed at agreed costs or at less cost.	[24]
	x17	SC: the project has not incurred all costs of taking appropriate measures to achieve the contract goal of schedule	[25]
	x18	QC: all necessary expenses and economic losses incurred in failure to meet quality standards	[25]
	x19	CC2: the project has not incurred additional costs due to the rate of requests for information	[7,12]

for details).

3. Research Methodology

3.1. Technical Route

Although some parts in this paper are analyzed qualitatively and quantitatively by summary and induction, this paper employs mainly the deductive logical analysis to conduct the study from the perspective of the full paper (e.g., establishment of quantified model). The technical route is shown in Figure 4. It can be decomposed into three layers: (1) Research foundation—The paper first proves the necessity of applicability of IPD principles via scientometric analysis and summarizes the research results that are directly related to the research topic through a literature inductive method. Then, it verifies that China is suitable for adopting IPD as a philosophy through the first questionnaire survey. Finally, it establishes the conceptual model for IPD principles application analysis via a qualitative analysis. (2) Questionnaire development—It focuses mainly on the second questionnaire survey, including the questionnaire design and the distribution, screening, and testing of raw data. Finally, this paper obtained two sets of sample data. (3) Data analysis—The structural equation model consists of a measurement model and a structural model. The measurement model can analyze the hypothesized relationship between latent variables and their observed variable, while the structural model links the independent and dependent latent variables [26]. However, this paper analyzes mainly the applicability of IPD principles, so it involves only the measurement model. The process of data analysis consists of establishing a measurement model. To verify the conceptual model, this paper first establishes two measurement models via questionnaire, and then analyses the results on the basis of the already established measurement model. Thus, this paper mainly conducts a construction validity test to determine whether the obtained results meet the requirements. As for the validity test, it includes mainly the factor analysis feasibility test, the exploratory and the confirmatory factor analysis (CFA), the convergent validity, and the discriminant validity. Moreover, this paper conducted a reliability test.

3.2. Questionnaire Development

In the following, the three main parts of the questionnaire are presented in detail.

3.2.1. Questionnaire Design

The majority of the second questionnaire’s content consists of basic informational statistics and model measuring scales (https://www.wjx.cn/jq/7855429.aspx (accessed on 31 October 2021)) Appendix A. Given the actual situation in the process of project management, the respondents should choose the most suitable option for most of the project (e.g., more inconsistent, slightly inconsistent, uncertain, slightly consistent, or more consistent). The second questionnaire used a matrix, including 5 options for the horizontal axis, the 15 IPD principles, and 4 indices of project performance as the longitudinal axis (shown in Table 1).

3.2.2. Questionnaire Distribution and Screening

The second questionnaire was distributed through conferences, e-mails, the online questionnaire platform, WeChat, and so on. A total of 310 questionnaires were received from 350 distributed questionnaires, so the questionnaire retrieving rate was 88.6%. Due to the low degree of application of IPD and the awareness of industry participants [9], most practitioners may not be familiar with IPD-related concepts, or they may even think that they are without relevant IPD experience. However, IPD principles, such as early involvement of participants, BIM, mutual respect and trust, willingness to collaborate, and open communication have been applied in other traditional delivery models in practical projects. To be more in line with the actual condition, such respondents will not be excluded. Thus, two sample groups have been selected: the average level of sample (ALS), for those who are inexperienced and unfamiliar with IPD, and the best level of sample (BLS), for those who are experienced with IPD or inexperienced but informed about IPD. After screening, the dataset of the ALS consisted of 205 participants, whereas that of BLS was formed of 105 participants. Since both sample sizes are greater than 100, it can be judged that the sample size essentially meets the requirements [27].

3.2.3. Test of Raw Data

The questionnaire is designed on a five-point Likert scale. There are no missing values and possible outliers in recollected questionnaires. The variables are normally distributed, according to the Q-Q graphs’ results. A chi-square test of the early and late respondents in the two samples showed that there does not exist any significant difference between them (p-value > 0.05) concerning the years of working and years of BIM use, while excluding the possibility of nonresponse bias. The variance of the first factor based on Principal Component Analysis (PCA) in the two samples accounted for 44.201% and 43.146% of the total variance of the factors, respectively, fulfilling the requirement of less than 50% [28]. Thus, the common source bias is not serious and is within the acceptable range.

3.3. Data Analysis

The two groups of sample data need to be conditionally tested via factor analysis before constructing the validity test in order to determine whether to conduct factor analysis. The factor analysis feasibility test of data sets is usually conducted via Kaiser–Meyer–Olkin (KMO) measurement of sampling adequacy and the Barlett’s test of sphericity. If the value of KMO is greater than 0.5, the data will be suitable for factor analysis [29]. Considering the cognitive status of Chinese practitioners and researchers on the IPD principles, this paper assumes that it is not yet determined the number of factors (the categories of IPD principle) of which the observed variable (IPD principles) consists nor the subordinate relationship between factors and observed variables. To make the conceptual model more realistic, the number of factors and the subordinate relationship between variables and factors will be determined by exploratory factor analysis (EFA). EFA uses the “dimensionality reduction” idea of PCA to eliminate the correlation between variables. This paper adopts two criteria: the eigenvalue cannot be less than or equal to 1, and the factor load cannot be less than 0.5. In addition, the variance maximum rotation method is used to orthogonally transform the initial factor load matrix and explain the potential variables more clearly. CFA is a special application of structural equation model. A measurement model will be established by using the CFA, evaluating how the observed variables will reflect the latent variables they intend to measure. At present, the academic world uses CFA mainly to test the construct validity, where the fitting indexes include the ratio of chi-square to degrees of freedom (χ²/df), the root mean square error of approximation (RMSEA), the non-normed fit index (NNFI), and the comparative fit index (CFI). After the factors are named, the use of CFA to test the results of EFA can further determine the description details of the conceptual model.

4. Results and Analysis

4.1. Data Descriptive Statistics

The 205 respondents come mainly from state-owned enterprises and private enterprises (they constitute approximately 78% of the total participants), followed by public institutions, whereas the proportion of other units is no more than 5%. Among the first category, state-owned enterprises account for 42%. The respondents are composed mainly of construction units, design institutes, and consulting units, with the proportions of 34.6%, 25.4%, and 13.7%, respectively. However, the owners represent just 8.8%, and the proportion of universities or research institutes is 9.8%. The supervisor department, considered as a unique by-product of the construction industry in China, has not paid attention to IPD, which may be due to the lack of reference of IPD projects by supervision engineers. In addition, 41.9% of the ALS sample and 38.04% of the BLS sample worked for more than five years. Among them, 70.3% of the first sample used BIM for less than 3 years (40%), 3~5 years (24.4%), and more than 5 years (5.9%), showing that BIM is becoming increasingly more popular in China, as most users have been using it for the past last years. For the respondents having more than 10 years of work experience, the proportion of using BIM technology for BLS is higher than that of ALS, indicating that the practitioners of BLS use relatively mature BIM technology.

4.2. Validity Test

4.2.1. Feasibility Test of Factor Analysis

If Bartlett’s sphericity test is significant, there is a strong correlation among variables [30]. As for KMO values, they are much higher than 0.50, indicating that the variables are interdependent (for the ALS, KMO = 0.939, Sig. = 0.000, whereas for the BLS, KMO = 0.907, Sig. = 0.000). Therefore, they are suitable for factor analysis.

4.2.2. Exploratory Factor Analysis

In this paper, a two-stage PCA method was used to further unfold the relationship among the observed variables. Considering ALS as an example, the three principal components eigenvalues were 9.160, 1.313, and 1.134, respectively, in the first PCA, and the final cumulative variance contribution rate was 61.087%. As for the second stage of PCA, the factor load was not less than 0.5, the principal component eigenvalues of the three principal components were 6.317, 0.823, and 0.658, respectively, while the final cumulative variance contribution rate was 70.883%. Considering BLS as an example, the results of the two-stage PCA are presented in

Table 2. Rotated component matrix of principal components analysis for the best level sample.

Principal Components Analysis	Observed Variable	Component
		1	2	3
The first principal components analysis	x1	0.782	0.232	0.141
	x2	0.754	0.305	0.259
	x3	0.738	0.157	0.194
	x4	0.821	0.217	0.179
	x5	0.556	0.320	0.396
	x6	0.537	0.169	0.541
	x7	0.500	0.550	0.290
	x8	0.439	0.369	0.350
	x9	0.665	0.259	0.284
	x10	0.606	0.405	0.154
	x11	0.587	0.415	0.252
	x12	0.153	0.702	0.293
	x13	0.266	0.796	0.104
	x14	0.333	0.782	0.163
	x15	0.350	0.542	0.215
	x16	0.192	0.362	0.687
	x17	0.133	0.338	0.767
	x18	0.222	0.489	0.594
	x19	0.389	0.066	0.737
The second principal components analysis	x1	0.624	0.398	0.284
	x2	0.721	0.316	0.370
	x3	0.713	0.472	−0.095
	x4	0.717	0.448	0.158
	x5	0.759	0.145	0.298
	x6	0.748	0.125	0.286
	x7	0.598	0.287	0.422
	x8	0.265	0.211	0.863
	x9	0.314	0.558	0.571
	x10	0.202	0.839	0.302
	x11	0.349	0.783	0.166

Note: Extraction Method—Principal Component Analysis. Rotation Method—Varimax with Kaiser Normalization. Rotation converged in 8 iterations. The bold of column indicates that the variables preceding the column belong to the same component.

. Last but not the least, the results of the EFA based on PCA show that 5 latent variables are needed to explain 19 observed variables in ALS and BLS, respectively.

4.2.3. Confirmatory Factor Analysis

Given the results of the two-stage PCA for the two groups of samples and the current situation of the use of the IPD principles in China, this study makes slight changes to the naming of the potential variables on the basis of the classification from NASFA et al. [11], AIA, and Minnesota [10]. The five latent variables for ALS and BLS were renamed as follows: project performance, contractual principles, catalysts, collaboration-related principles, and behavioral principles (as shown in

Table 3. Renaming of factors after exploratory factor analysis.

Conceptual Model			ALS			BLS
Factor	Observed Variable	Item/Indicator	Factor	Observed Variable	Item/Indicator	Factor	Observed Variable	Item/Indicator
Contractual principles	x1	KPBTE	Contractual principles (CP1)	x1	KPBTE	Contractual principles (CP2)	x1	KPBTE
	x2	LWKP		x2	LWKP		x2	LWKP
	x3	EIKP		x6	SRR		x3	EIKP
	x4	FT		x4	FT		x4	FT
	x5	JDPTC		x7	ID		x5	JDPTC
	x6	SRR	Collaboration-related principles (CRP1)	x3	EIKP		x6	SRR
	x7	ID		x5	JDPTC		x7	ID
	x8	CDM	Behavioral principles (BP1)	x8	CDM	Collaboration-related principles (CRP2)	x8	CDM
Behavioral principles	x9	MRT		x9	MRT		x9	MRT
	x10	WTC		x10	WTC	Behavioral principles (BP2)	x10	WTC
	x11	OC		x11	OC		x11	OC
Catalysts	x12	MA	Catalysts (C1)	x12	MA	Catalysts (C2)	x12	MA
	x13	BIM		x13	BIM		x13	BIM
	x14	LC		x14	LC		x14	LC
	x15	CT		x15	CT		x15	CT
Project performance	x16	CC1	Project performance (PP1)	x16	CC1	Project performance (PP2)	x16	CC1
	x17	SC		x17	SC		x17	SC
	x18	QC		x18	QC		x18	QC
	x19	CC2		x19	CC2		x19	CC2

). The collaboration-related principles can be applied to the maximum extent if the parties agree to collaborate. The results of EFA of the two groups of samples are essentially consistent with the composition of the conceptual model. Hence, some differences are identified as follows: (1) respondents representing the average level sample believe that collaborative decision-making is a behavioral principle, and they further subdivide the original contract principle into two categories; and (2) respondents representing the best level sample believe that collaborative decision-making and mutual respect and trust belong to the collaboration-related principles rather than the behavioral principles or contractual principles. As for the main reasons for these differences, they can be summarized by the following: (1) collaborative decision-making can be regarded as the result of multiparty contract which can support foreign IPD projects; and (2) the multiparty contract is difficult to achieve in China at present. Thus, a collaborative decision-making has been perceived as an owner-driven, spontaneous behavior characteristic of participants rather than being a mandatory clause in a contract.

This paper compares the difference between the regenerative covariance matrix and the sample covariance matrix to test whether the model fits the data. The fitting index is used to represent the overall difference between the two matrices. By eliminating the observed variables with loadings (standard coefficient) on potential variables lower than 0.6, two groups of modified measurement models are obtained on the basis of multiple iterations (for ALS, χ²/df = 1.917, RMSEA = 0.067, NNFI = 0.980, and CFI = 0.980; for BLS, χ²/df = 0.594, RMSEA = 0.080, NNFI = 0.97, and CFI = 0.97), achieving the threshold value of each fit index recommended in the references (χ²/df < 3 [31], RMSEA < 0.08 [32,33], NNFI ≥ 0.90 [33], and CFI ≥ 0.90 [34]). Figure 5 shows the modified measurement model. It can be concluded that the deleted variables are the open communication and the co-location of team for the ALS sample, whereas it is the intensified design variable for BLS sample.

4.2.4. Convergent Validity and Discriminant Validity

By comparing observed data, the measurement model assesses how effectively underlying hypothesized constructions may be identified. When standardized factor loadings exceed 0.50, it has a high convergence in statistical significance [32]. The value of average variance extracted (AVE) and composite reliability (CR) will be greater than or equal to 0.5 and 0.7, respectively; both requirements should be met for convergence validity. For ALS, the values of CR of CP1, CRP1, BP1, C1, and PP1 are, respectively, 0.887, 0.732, 0.846, 0.807, and 0.794, while the values of AVE are, respectively, 0.642, 0.578, 0.579, 0.587, and 0.563. As for BLS, the values of CR of CP2, CRP2, BP2, C2, and PP2 are, respectively, 0.901, 0.827, 0.777, 0.820, and 0.815, while the values of AVE are, respectively, 0.604, 0.706, 0.638, 0.539, and 0.525. Hence, the convergent validity of the constructs in the two measurement models meet the requirements. To confirm the discriminant validity, the inter-correlation between the constructs should be less than the suggested threshold of 0.85 [32]. Figure 5 presents that the discriminant validity of the measurement model at BLS is more ideal than that at ALS.

4.3. Reliability Test

Cronbach’s alpha is the most commonly used estimate of internal consistency. It is often used to evaluate the correlation among the simultaneous project responses. When alpha value is greater than 0.7, the reliability of the scale is high. For ALS, the values of alpha for CP1, BP1, CRP1, C1, and PP1 are, respectively, 0.885, 0.832, 0.733, 0.79, and 0.799, whereas for BLS, the alpha values for CP2, BP2, CRP2, C2, and PP2 are, respectively, 0.899, 0.827, 0.771, 0.815, and 0.813. The variables for both samples have alpha greater than 0.7; thus, the reliability level is high. The overall Cronbach’s alpha at ALS and BLS are 0.939 and 0.943, respectively. Thus, the internal consistency coefficient of the calculation scale of all variables essentially meets the requirements, and the observable variables of each sample are relatively reliable.

5. Discussion

Through the comparative analysis for the two sets of measurement models, this paper clearly reveals the internal relationships among each IPD principle in the application process. On the basis of the measurement models, this paper provides a scientific and reasonable basis for how to effectively apply the IPD principles.

As we are concerned mainly with the applicability of IPD principles, the discussion involves primarily the qualitative analysis of the effectiveness of the application of IPD principles, aimed at how to guide practical projects to implement IPD. It is reflected in two main aspects: first, the order of importance of each IPD principle in each category can be determined on the basis of the loading factor of the measurement model, and then the principles with higher importance can be used preferentially in practical projects; and second, the correlation coefficient among each category of IPD principle and project performance and other categories of IPD principles is used to understand the degree of correlation, and then it is demonstrated that these principles need to be combined in practical projects. According to Figure 4, the application status and the effect of IPD principles will vary due to the presence of different users. The detailed discussion follows.

5.1. Project Performance

Different users emphasize different measurement indicators for the project performance. The loading factor of construction cost, schedule cost, and quality cost on the project performance in ALS are 0.78, 0.75, and 0.72, respectively, while the loading factor of construction cost, schedule cost, quality cost, and communication cost on project performance in BLS are 0.73, 0.74, 0.79, and 0.63, respectively. It can be seen that the respondents who are inexperienced and unfamiliar with IPD think that the construction cost is the most important, and the communication cost is not considered when measuring the project performance. That is also why the respondents who are experienced with IPD or inexperienced though informed about IPD hold that the most important factor in the project performance is the quality cost, and they pay much attention to the communication cost. The highly integrated project delivery method facilitates quick, accurate, and comprehensive sharing of the opinions of the project professionals, reducing the deviations and delays associated with information transmission [35]. Therefore, the use of IPD principles will help in improving communication efficiency [18].

All respondents agreed that the project performance is related to IPD principles. The correlation coefficient between project performance and collaboration-related principles is the largest (equal to 0.76) for those who are inexperienced and unfamiliar with IPD, while the correlation with the catalyst is the largest (equal to 0.74) for those who are experienced with IPD or inexperienced though informed about IPD. Some research proved that the IPD principles can also improve project results [36]. With an increase in the number and intensity of collaboration and integration principles and methods, project performance will improve [37]. The collaboration and the integration principles and methods required early involvement of key participants, jointly developed project target criteria, collaborative decision-making, shared financial risks and rewards based on project outcome, co-location of team, liability waivers between key participants waivers, multiparty agreements, intensified design, etc. Thus, the application of IPD principles is very indispensable to improving the project performance or its results.

5.2. Collaboration-Related Principles

The early involvement of key players and jointly developed project target criteria need to be guaranteed in the contract, while collaborative decision-making and mutual respect and trust cannot just be patterns of behavior. Thus, one can identify the following:

(1)

The respondents who are inexperienced and unfamiliar with IPD consider that the latter two principles (e.g., early involvement of key participants and jointly developed project target criteria) belong to the collaboration-related principles and are the most relevant to the contractual principles (correlation coefficient is equal to 0.88). In addition, IPD principles require all parties to contribute knowledge to the common goal at the early stage of the project [38]. The main participants’ early involvement and working on mutual trust and benefits display that IPD has significant advantages in reducing design changes [15]. The owners must encourage participants to choose collaborative behavior [39], as they may consider the impact of early intervention of key participants and joint formulation of project objectives and standards when signing the contract. This coincides with the idea of Breyer et al. [40], who recommended integrating builders early in the design process, binding economic interests with mutual project goals, and collaborating with participants in the design and construction process.

(2)

The respondents who are experienced with IPD or inexperienced though informed about IPD insist that the collaboration-related principles, including collaborative decision-making and mutual respect and trust, have the greatest correlation with the behavioral principles (correlation coefficient: 0.77). Moreover, mutual respect and trust (loading factor: 0.85) is more essential than collaborative decision-making (loading factor: 0.83). IPD will overcome the obstacles of cooperation and increase the early engagement of key participants and the trust among key stakeholders [5]. Mutual respect and trust, good leadership, strong team involvement, and effective working relationships are the main motivations for applying IPD. Among them, trust is the main driving force to promote collaboration among stakeholders of different projects [41]. Thus, they may believe that IPD may focus more on collaborative decision-making and mutual respect and trust, rather than limiting them to a simple behavior pattern.

5.3. Behavioral Principles

Despite the difference in the classification of IPD principles, the interaction between behavioral principles and contractual principles will affect project performance. This can be modeled at different levels:

(1)

The respondents who are experienced with IPD or inexperienced though informed about IPD believe that the behavioral principles mainly include open communication and willingness to collaborate, among which open communication has the greatest impact on behavioral principles (load coefficient: 0.87). The commitment levels of team members are closely related to the behavioral principles. Only when team members have high and uniform commitment is it easy to avoid issues such as goal dislocation, poor communication, and lower decision quality [42]. Moreover, the reduction of project cost overrun is affected mainly by open communication, whereas the reduction of on-site rework costs is significantly affected by collaborative decision-making [2]. The future willingness to collaborate can effectively encourage cooperative behavior and reduce risks in the projects [43].

(2)

The respondents who are inexperienced and unfamiliar with IPD believe, at all levels, that the impact of the behavioral principles is mutual respect and trust (path coefficient: 0.79), willingness to collaborate (path coefficient: 0.78), open communication (path coefficient: 0.77), and collaborative decision-making (path coefficient: 0.70). An important element of team cohesion is full communication and mutual trust [43], where trust is the core mechanism that can improve cooperation among IPD project parties [38]. The first three main principles create a collaborative environment, while the latter principle emphasizes decision-making behavior in such an environment. Project vision, behaviors of project participants, communication, and contractual agreement are the important factors that affect the cooperation of construction projects [44]. Therefore, all respondents consider that behavioral principles are the most relevant to the contractual principles, proving that the interaction among them will have an impact on project performance.

5.4. Catalyst

The catalyst of IPD has a great correlation with the project performance, among which lean design and construction as well as BIM technology are the main catalysts. The respondents who are inexperienced and unfamiliar with IPD believe the importance ranking of all catalysts is related to the lean design and construction (load factor: 0.86), BIM technology (load factor: 0.79), and multiparty agreement (load factor: 0.63). However, the respondents who are experienced with IPD or inexperienced though informed about IPD consider the importance ranking of all catalysts is related to the lean design and construction (load factor: 0.86), BIM technology (load factor: 0.82), multiparty agreement (load factor: 0.63), and co-location of team (load factor: 0.59). The corresponding time of RFIs can be greatly shortened through co-location of teams [2]. Compared with the latter respondents, the former respondents have not realized the direct impact of co-location of team in project management. The relevant domestic literature shows that some scholars have described conceptual co-location or mentioned it only in some engineering phases. However, all respondents agreed that BIM technology and lean design and construction were the main catalysts for the application of IPD principles.

Lean project delivery system is one of the effective ways to improve productivity, especially in achieving lean construction principles. IPD will help to remove obstacles to the application of BIM [5], while BIM is the technical support for the successful implementation of IPD [45]. Moreover, the integration of BIM and IPD can improve overall results of the design and construction processes [46]. Thus, the most important solution to enhancing collaboration is the integrated application of BIM and IPD [47]. However, the cornerstone of realizing IPD targets is multiparty agreement among key participants [48]. The latter respondents believe that the catalyst has the greatest correlation factor with project performance (correlation coefficient: 0.74), while the former respondents insist that it has a greater correlation with behavior principles (correlation coefficient: 0.73). Therefore, it can be judged that the catalysts will affect project performance.

5.5. Contractual Principles

Contractual principles are the main bridge to ensure the interaction of other IPD principles. Practitioners and researchers have theoretically recognized the importance of liability waivers and fiscal transparency among key participants, as well as shared financial risk and reward based on project outcome. All respondents believed that key participants bound together as equals and shared financial risk and reward based on project outcome are equally important for contractual principles. Respondents, who are experienced with IPD or inexperienced though informed about IPD focus more on important players who are bonded as equals and mutually defined project aim criteria but neglect to take intensive design into account. The basic step to achieving mutual trust and open communication is to establish equality and mutual respect among project team members [49]. The alignment of the project target, the shared financial risk and reward based on project outcome engenders a spirit of cooperation and teamwork. Although some IPD principles cannot be directly achieved in China, such as fiscal transparency among key participants and shared financial risk and reward based on project outcomes, all respondents still believe those principles very important in the contract principles. Moreover, the respondents who are inexperienced and unfamiliar with IPD think that the contractual principles have the highest correlation with the principle of coordination (correlation coefficient: 0.88), while the respondents who are experienced with IPD or inexperienced though informed about IPD hold that the highest correlation exists the behavioral principles (correlation coefficient: 0.80). The IPD approach, based on relational–contractual principles, depends on cultivating trust and commitment of team members. By sharing project goals in the early stage of project delivery, they can collaborate and resolve project-related issues and share risks and rewards [50]. Therefore, the contractual principles may be the main bridge along with other IPD principles.

6. Conclusions

This paper provides a way to clarify the applicability of the IPD principles in a certain region or country which has a specific cultural background. Considering China as an example, the measurement model of applicability analysis of IPD is established over two groups of sample data. This research broadens our understanding of project delivery systems.

First, on the basis of qualitative and quantitative analysis, the paper proposes a new idea to study the applicability of IPD principles which makes up the previous research on IPD application. (1) The application analysis of IPD principles is very meaningful, but the directly related research is very limited. Previous studies advocated the mutual application of IPD principles and other technologies and demonstrated that the application of IPD principles has an impact on project performance or results. (2) The applicability analysis of IPD principles lays a theoretical foundation for countries and regions which use IPD as a philosophy, aiming to promote the development of BIM technology. (3) China is suitable for using IPD as a philosophy rather than a delivery method at present. However, the owners’ enthusiasm for embedding the integrated application of BIM and IPD is not as high as that of universities or research institutes, design institutes, and even contractors. With the gradual popularization of BIM technology in China, most IPD principles have been applied in practical engineering projects by the state-owned enterprises, even if some IPD principles should be hindered in China. (4) Considering the construction cost, schedule cost, quality cost, communication cost, and the 15 IPD principles as observation variables, a measurement model for the applicability analysis of IPD principles is constructed. The measurement models of IPD applicability analysis of the two sample groups are essentially consistent with the composition of the conceptual model.

Second, the internal influence relationship among IPD principles will lay a theoretical foundation for subsequent research regarding the influence mechanism of IPD principles on project performance. Thus, one can recognize the following: (1) On the basis of China’s cultural background, this paper classifies IPD principles and reveals the internal influence relationship among IPD principles with the help of a measurement model. The importance ranking of each principle in each classification of IPD principle reflects the priority of IPD principle in practical engineering, providing a scientific reference for the efficient application in other countries or regions where IPD is deemed as a philosophy. (2) The project performance will be promoted through the targeted application of IPD principles in considering the different regions and users. Different users pay various levels of attention to the measurement indicators of project performance. However, all measurement models consist of contractual principles, behavioral principles, collaboration-related principles, as well as catalysts for IPD. The latter of these is closely related to project performance. All these factors lead to the conclusion that IPD can affect project performance, as obtained by previous scholars.

Finally, there are still some limitations: (1) The results of this study are based on questionnaire data in China; future work may extend this study to other countries with different culture backgrounds. (2) Due to the significant differences in the development of BIM technology and the cognitive level of IPD in different regions in China, the regional differences will affect the implementation of IPD principles. The practitioners’ understanding of collaborative work, the needs of construction projects, the project management level, etc., will also propose regional discrepancy in the context of worldwide construction industry. (3) This paper conducts mainly a qualitative analysis of the effectiveness of the application of IPD principles; subsequent research may attempt to use different quantitative analysis methods to deeply measure the effectiveness of the application of IPD principles. (4) Considering the role of different cultural factors and considering these cultural factors as conditions and the application effect of IPD principles as results, the future research will further reveal the configuration path of various cultural factors affecting the application effect of IPD principles by using the QCA method. (5) This paper conducts only a quantitative analysis of the internal relationship between IPD principles, further research needs to deeply reveal the impact mechanism between IPD principles and project performance for truly defining how to scientifically and rationally use IPD principles to improve project performance.