A Synthesis of Express Analytic Hierarchy Process (EAHP) and Partial Least Squares-Structural Equations Modeling (PLS-SEM) for Sustainable Construction and Demolition Waste Management Assessment: The Case of Malaysia

Abstract

Construction and demolition waste (CDW), as the main consequence of construction and demolition (C&D) activities, has severely affected our sustainability needs. However, construction and demolition waste management (CDWM) lacks the integration of sustainability concepts. Thus, there is a great need to include sustainability dimensions in CDWM to reach sustainable construction and demolition waste management (SCDWM). This study aims at empirically investigating SCDWM by analyzing the impacts of factors that contribute to sustainability aspects of CDWM on waste management hierarchy (WMH), including reduce, reuse, recycle, and disposal strategies. According to the literature, 26 factors were initially identified and grouped under four categories, namely environmental, economic, social, and administrative, that contribute to sustainability aspects of CDWM. Subsequently, a pilot test was performed to assess the significance and applicability of these factors in the Malaysian construction industry by implementing the express analytic hierarchy process (EAHP). Then, a questionnaire survey was performed to collect data from 132 construction companies involved in CDWM. Partial least squares-structural equation modeling (PLS-SEM) was used to test the hypothetical relationships by applying SmartPLS software. Results demonstrated that the economic aspect of CDWM (main category) and “public environment contamination due to illegal waste dumping” (sub-category) were the most influential factor in SCDWM in Malaysia.

1. Introduction

Economic development, along with population growth, call for massive construction activities worldwide [1]. Consequently, a massive volume of construction and demolition waste (CDW) is engendered, which is the principal cause of harmful impacts on the environment, society, and the economy [2,3,4]. In this case, natural resource depletion, air and water pollution, costs linked to waste collection, sorting, recycling, and disposal, and public safety and health concerns are among the impacts that can be attributed to CDW [5]. The estimated quantity of globally produced CDW each year is over 10 billion tonnes [6,7], of which almost 9.5 million tonnes is produced in Malaysia [8]. Excessive CDW generation, along with the low recycling rate of almost 15% in Malaysia [9] have prompted the need for a sustainable management mechanism for CDW, in which economic benefits, social impacts, and environmental control are fulfilled [10,11,12]. Thus, sustainable construction and demolition waste management (SCDWM) is taken into consideration, which encompasses reducing raw material consumption, reusing materials, appropriate recycling mechanisms, and minimizing waste generation from construction and demolition (C&D) tasks [10,13,14,15].

The concept of sustainability in construction and demolition waste management (CDWM) is mostly attributed to the strategies or factors that influence CDWM sustainably. For instance, [16,17] pointed out the most acceptable and commonly appropriate strategy for managing CDW sustainably by employing 3Rs embracing reduce, reuse, and recycle strategies. However, when CDW cannot be treated through reducing, reusing, and recycling, safe disposal of waste is implemented [18,19]. These strategies are predominantly referred to as the waste management hierarchy (WMH) [20]. It is worth considering that although this hierarchy includes other strategies, such as energy recover, these strategies are not common for CDWM in its general form. In addition to strategies, there are also factors that have been proven influential in managing CDW sustainably. For instance, Tsai, et al. [21] pointed out the triple bottom lines (TBL) of sustainability with respect to the factors that influence CDWM from environmental, economic, and social outlooks, such as excessive waste recycling costs [22], soil, air and water pollution [1], public health hazards [23], etc.

However, CDWM lacks the integration of sustainability concepts [24]. Also, the traditional practices used for SCDWM in the majority of developing countries, such as Malaysia, fall behind those of developed countries e.g., [25,26,27]. Thus, it is essential to take sustainability dimensions of CDWM into account in order to attain SCDWM. Therefore, this study aims at assessing SCDWM. To do so, first, determinants (factors and strategies) that contribute to SCDWM are identified. Then, a conceptual framework capable of analysing SCDWM is developed, in which determinants of SCDWM are juxtaposed in a logical way to from hypotheses with which the logic of assessment is fulfilled. Finally, with the help of empirical data from Malaysian construction professionals, the effectiveness of SCDWM is assessed.

2. Literature Review

The increasing rate of construction, renovation, and demolition activities around the world has produced a massive amount of CDW [17,28], which primarily consists of metals, wood, plastics, cardboards, plasterboard, asphalt, bricks, concrete, glass, and other materials [29]. It may also contain substances that adversely impact human health as well as the natural environment. Additionally, CDW contributes to economic loss and natural resource depletion [4]. Thus, CDW in its whole life cycle including generation, collection, transportation, recycling, and disposal ought to be managed sustainably [30] (otherwise it threatens our natural environment) [14]. In order to accomplish SCDWM, it is prominent to meticulously review strategies and factors that influence CDWM from sustainability perspective.

2.1. Sustainability Concept in CDWM

The incorporation of three pillars of sustainability embracing social, economic, and environmental activities have primarily been highlighted by sustainable development (SD) [31]. When SD is integrated with the construction industry, sustainable construction (SC) emerges, which refers to generating a sustainable and healthy environment in which buildings and infrastructures have a beneficial impact on natural systems when seen through a balanced economic, environmental, and social perspective [32]. Similarly, when SD and SC are incorporated with waste management, sustainable waste management (SWM) is initiated, which refers to objectives of protecting environment and health while conserving natural resources [7,33]. SWM was then altered to the waste management hierarchy (WMH), the principal primacy in CDWM [34]. In other words, the SWM was evolved later inside the WMH, which is considered the most applicable strategy for managing CDW [16,35,36]. WMH includes reduce, reuse, recycle (3Rs) strategies, and safely dispose of waste in order of preference from the most to the least favorable strategies. Thus, SCDWM could be described as balancing the sustainability aspects of CDWM and WMH on the other. The following sections (Section 2.2, Section 2.3, Section 2.4 and Section 2.5) explain sustainability aspects of CDWM.

2.2. Environmental Aspect of CDWM

Inappropriate management of CDW poses harmful impacts on the environment by taking up land resources as landfills [30,37,38], generating harmful pollutants that endanger the atmosphere, and taking over natural resources that are already limited [14,39,40]. Landfills, for instance, are accountable for large amounts of methane gas emissions via decomposition of mixed CDW overtime, which is more harmful than CO2 with regards to global warming impacts [30]. Several researchers have also addressed the environmental concerns associated with CDW, such as soil degradation [17,41,42], water pollution [43], and global warming [44,45]. According to Ding, et al. [46], every square metre of landfilled CDW results in 53 g of soil deterioration and about 1.5 tonnes of groundwater loss. Other researchers have also pointed out the noise and environmental pollution associated with legal/illegal CDW collection, transportation and disposal caused by construction machineries [47,48].

2.3. Economic Aspect of CDWM

The economic consequences of CDW are particularly substantial since construction materials account for almost 40% of the material stream within the global economy [49,50]. Additionally, material costs account for 50%–60% of the total cost of a construction project [51]. However, approximately, 20–30% of construction materials are wasted during the construction stage and are sent to landfills [52]. The economic aspect of CDWM mainly includes costs related to material use, tax on landfills, loss of productivity, disposal costs, transportation and overhead costs, recycling and reuse costs, operating costs of recycling facilities, etc. [37,53,54,55]. According to Lockrey, et al. [56], economic viability is an essential factor that substantially affects the performance and behavior of CDWM practitioners. Additionally, Tura, et al. [57] highlighted that the primary impediment to engaging SCDWM practices is a high level of economic uncertainty; for example, the processes involved in recycling are costly compared to the value of the recycled products. Furthermore, Chen et al. [58] came up with the economic concerns as the most influential factor in CDWM from both governmental and institutional perspectives. Besides, Yuan [59] emphasized the role of economic concerns towards the slow and ineffective process of waste recycling at recycling facilities.

2.4. Social Aspect of CDWM

A significant obstacle to the development and implementation of SCDWM practices is the presumption that these practices result in the augmentation of project costs [10,24]. Lockrey, Nguyen, Crossin and Verghese [56] highlighted the role of contractors’ limited knowledge, awareness, and inadequate community involvement as the main barriers to implementing SCDWM practices. Similarly, Abarca-Guerrero, et al. [60] asserted that the social awareness is an effective factor in achieving SCDWM. In this regard, when CDWM stakeholders believe that waste generation is necessary, implementing sustainable strategies is ineffective and challenging. Likewise, Blaisi (2019) [37] came up with the lack of commitment, inadequate collaboration, and poor perception among CDWM practitioners as challenges that impede CDWM from achieving sustainability from social perspective. Additionally, safety and health issues related to the materials’ recycling from collection to transportation and recycling, public awareness, aspiration, and satisfaction towards CDWM, decisions and attitudes of CDW stakeholders, public participation in planning CDWM practices, and social concerns with illegal dumping of CDWM have been addressed by some scholars e.g., [61].

2.5. Administrative Aspect of CDWM

Along with the environmental, economic, and social aspects of sustainability (TBL) in CDWM, administrative concerns have also received attention in recent years. Administrative concerns associated with CDWM, include a wide range of factors. For instance, the availability/unavailability of recycling facilities can significantly impact waste management as it promotes the illegal dumping of waste [62,63]. Similarly, difficulties such as technical hurdles, lack of expert with substantial knowledge, and experience in waste management, and modern technology often impede the implementation of sustainable waste management strategies [21]. Also, Mahpour [24] pointed out the technical and legal issues as substantial barriers in obtaining sustainability and circular economy in waste management. Insufficient supervision and weak policies have also been addressed as major obstacles in achieving sustainability in the construction industry [64,65]. Besides, the role of governance in CDWM has been highlighted in the work of many researchers in the form of administrative, including regulations, technical considerations, management, and responsibility issues [10,59,66,67].

This study considers sustainability in CDWM from environmental, economic, social, and administrative perspectives.

Table 1. Contributing factors to sustainability aspects of CDWM.

Category	Code	Factors	References
1. Environmental	ENP1	1.1. Land consumption due to waste	[10,23,30,34,68]
	ENP2	1.2. Noise pollution
	ENP3	1.3. Water pollution
	ENP4	1.4. Air pollution
	ENP5	1.5. Public environment contamination due to illegal waste dumping
2. Economic	ECP1	2.1. Waste sorting, collection, and separation costs	[30,37,54,69]
	ECP2	2.2. Costs associated with reusing waste
	ECP3	2.3. Costs associated with recycling waste
	ECP4	2.4. Costs associated with waste transportation
	ECP5	2.5. Costs associated with waste disposal
	ECP6	2.6. Penalty for illegal dumping of waste
	ECP7	2.7. Revenue from selling recycled materials
	ECP8	2.8. Saving in project costs
	ECP9	2.9. Incentive mechanism for waste management
3. Social	SOP1	3.1. Waste management practitioners’ awareness	[30,70,71]
	SOP2	3.2. Provision of job opportunities
	SOP3	3.3. Arduous physical working condition
	SOP4	3.4. Health related long impacts
	SOP5	3.5. Waste management operatives’ safety
	SOP6	3.6. Public satisfaction about waste management practices
	SOP7	3.7. Public demand for regulating illegal waste dumping
	SOP8	3.8. Aesthetic impacts of illegal waste dumping
4. Administrative	ADP1	4.1. Policy quality	[30,72,73]
	ADP2	4.2. Technology
	ADP3	4.3. Management and supervision
	ADP4	4.4. Responsibility issues

summarizes the factors that contribute to CDWM from sustainability perspective.

3. Conceptual Framework

This step entails the development of a conceptual framework for analysing SCDWM. The concomitance of pillars to sustainability concept of CDWM comprising economic, environmental, social, and administrative, and WMH, strategies through which waste is managed, shaped this research’s conceptual framework. In other words, four pillars of sustainability concept of CDWM including environmental, economic, social, and administrative concerns are hypothesized to affect WMH, including reduce, reuse, recycle, and disposal strategies. The developed framework contains four hypotheses through which the effects of contributing factors to sustainability aspects of CDWM on WMH are measured [74,75]. Figure 1 represents the conceptual framework of this study. Hypotheses are explained as follows:

The environmental aspect of CDWM has a direct impact on WMH: the literature indicated that the environmental aspect of CDWM plays an essential role in managing CDW, which is in accordance with Sev [76] who believed that future generations deserve a better world by protecting natural systems and ecological balance from destruction by reducing the amount of generated waste. Also, Hu [77] and Yates and Castro-Lacouture [78] emphasized the pivotal role of the environmental aspect considering the reusability and recyclability of CDW. Therefore, in line with the existing literature, it was hypothesized that the environmental aspect of CDWM influences WMH (H₁).

The economic aspect of CDWM has a direct impact on WMH: the related literature indicated that the economic aspect of CDWM plays an essential role in managing CDW, which is aligned with studies performed by [53,79]. Other scholars have also pointed out the factors, such as costs involved in reusing, reducing, recycling, and disposing of waste throughout waste collection, transportation, and disposal [80]. Therefore, in line with the existing literature, it was hypothesized that the economic aspect of CDWM influences WMH (H₂).

The social aspect of CDWM has a direct impact on WMH: waste management practices impose a series of social concerns such as health and safety associated with waste production, sorting, collection, recycling, reuse, transporting, and disposals [30]. The relationship between social aspect of CDWM and sustainability were particularly in line with Santos, Mendes and Ribau Teixeira [70] who acknowledged that the construction projects are affected by the users who engaged in projects. This was also in line with Almahmoud and Doloi Hemanta [81] who proved that the satisfaction level of different stakeholders is crucial to achieving social sustainability in projects. Therefore, in line with the existing literature, it was hypothesized that the social aspect of CDWM influences WMH (H₃).

The administrative aspect of CDWM has a direct impact on WMH: the relationship between the administrative aspect of CDWM and WMH was in line with Santos, Mendes and Ribau Teixeira [70] in the context of public waste management including training and research that provide knowledge on further management options for taking the responsibility to determine the role of stakeholders. The applied technologies for reducing and recycling waste are also significant [82]. Furthermore, the role of policies and supervision have been proven effective in reducing and reusing CDW [83]. Therefore, in line with the existing literature, it was hypothesized that the administrative aspect of CDWM influences WMH (H₄).

4. Materials and Methods

The effects of factors that contribute to the sustainability perspective of CDWM (factors) on WMH (strategies) should be empirically investigated to outline the principals of SCDWM in the Malaysian construction industry. The research process is illustrated in Figure 2. Empirical analysis could be performed with the help of quantitative data [84], and thus primary data needed to be collected as data pertinent to SCDWM concerning the developed framework could not be provided from secondary sources (e.g., government records). In this regard, the survey research method is definitely one of the most common methods of data collection in the areas of construction management and built environment [85]. Besides, the survey research method works well with the deductive research approach of this study, where research begins with reviewing other researchers’ works (general perspective) and continued with hypotheses formulation and testing (specific perspective) [86]. Additionally, the survey research method is effective in terms of collecting data from the target population in a reasonable timeframe [87].

Effective data collection along with providing less biased results compared to other methods (i.e., interviews) could be obtained using a questionnaire survey. Therefore, this method was used in this research [88,89]. Questionnaire surveys let researchers gather data from experts in a specific field and generalising the output to the public [89]. It is worth mentioning that the suitability and practicality of the research model including factors that contribute to sustainability perspective of CDWM was initially assessed through pilot testing using the express analytic hierarchy process (EAHP) (round 1) (Section 4.1). After performing pilot test (EAHP) and refining the factors through experts’ judgements, the main questionnaire survey (round 2) was distributed, data were collected, and then analyzed through partial least square based structural equation modeling (PLS-SEM) using SmartPLS version 3.0. Section 4.2 (SmartPLS GmbH, Bönningstedt, Germany). explains the process of data collection and analysis using PLS-SEM in detail.

4.1. Pilot Testing Using Express Analytic Hierarchy Process (EAHP)

AHP refers to an structured technique for analysing multi-criteria decision making (MCDM) problems according to a pairwise comparison scale [90], which manages a large number of decision factors and provides a taxonomic process to rank a lot of decision variables [24,91]. AHP has extensively been applied in different areas of construction management such as construction technologies assessment, material assessment, market decision making, contractors’ selection, equipment selection, etc. [92]. AHP as a subjective model without needing a large sample, is a proper method of focusing on a specific issue and prioritising the category/factor involved in any decision model [93]. However, AHP has some limitations, especially when a large number of category/factor are involved i.e., it requires excessive number of pairwise comparisons, and it is very likely to end up with inconsistency among the responses [92]. To tackle these limitations, in this study, a simplified version of AHP was employed, which was introduced by Leal [94] and called express AHP (EAHP), in which a very low number of pairwise comparisons were required compared to conventional AHP developed by Saaty [95].

The EAHP survey (round 1) was initially implemented in this study to determine the significance of contributing factors to the sustainability perspectives of CDWM in the Malaysian construction industry. The EAHP questionnaire survey (Appendix A) contained two sections: section one included demographic information encompassing the respondents’ gender, age, years of experience in managing CDW, academic qualification, and occupation, while the second section included pair-wise comparison questions on the economic, environmental, social, and administrative aspects of CDWM (Table 1). The application of the EAHP survey in this study fulfilled the pilot testing criteria as the categories and factors were meticulously evaluated and then refined by experts’ judgments. This step could be considered a prologue to the main questionnaire design and distribution [96]. In this study, ten professionals with extensive waste management experience in the Malaysian construction industry, who were willing to participate in the EAHP questionnaire survey, were selected. These included four experts from the Construction Industry Development Board (CIDB), three from the Solid Waste and Public Cleansing Management Corporation (SWCorp), one from a waste management contractor, one from an environmental consultant, and one from academia. Respondents were all male, over than ten years of waste management experience, holder of an advanced degree in the area of construction and engineering, and actively participated in professionally related forums, conferences, and workshops. To refine the factors, the following steps were undertaken:

Step 1. Develop the hierarchy tree. In this study, the hierarchy included the goal, categories, and factors, where the importance of all categories was compared pairwise against the goal, while the importance of all factors was compared pairwise against their respective categories. The categories and factors have been extracted by meticulously reviewing the relevant literature (Table 1).

Step 2. Select the most important category/factor (one factor only). Each expert was asked to identify the most important category and factor in each category. In case there were more than one important category/factor, which may have the same importance based on the expert’s opinion, the expert was asked to select any of them at this stage. According to Saaty and Özdemir [97], a panel consisting of at least seven experts is the most suitable size to adopt the EAHP method.

Step 3. Create a pairwise comparison matrix as follows.

$$A_{1\times n}={\left[a_{ij}\right]}_{1\times n}, j=1,\dots ,n$$

(1)

where, $i$ is the most important category/factor based on expert’s opinion, $n$ is the number of categories/factors, and $a_{ij}$ is the comparison value of category/factor $i$ with category/factor $j$. It is notable to mention that in this research 9-point scale is used to enhance the accuracy of the findings. The scale of pairwise comparison is shown in

Table 2. EAHP pairwise comparison scale.

Intensity of Importance *	Linguistic Value
1	Equally important
2	Equally to moderately more important
3	Moderately more important
4	Moderately to strongly more important
5	Strongly more important
6	Strongly to very strongly more important
7	Very strongly more important
8	Very strongly to extremely more important
9	Extremely more important

* Reciprocals are not used as all categories/factors are merely compared with the most important category/factor.

.

Step 4. Calculate the weight of each category/factor using Equation (2).

$$W_j=\frac{1}{a_{ij}\sum \frac{1}{a_{ik}}}, k=1,\dots ,n$$

(2)

The weight of factors (local weight) obtained using Equation (2) is multiplied by the weight of their respective category to obtain the global weight (GW) of each factor, which is the basis for the refinement of the factors. The average weights, obtained by analyzing the inputs from all experts, of all categories and factors were considered final global weights (FGWs). It is worth mentioning that since in EAHP, the pairwise comparison is carried out only among the most important category/factor with other categories and factors, there will be no matter of inconsistency and all responses were considered consistent. At this stage, the experts could provide their recommendations for refining the factors based on their final GW.

4.2. Main Survey Distribution (Round 2)

To analyse the determinants of sustainability aspects of CDWM on WMH, the second survey (main survey) was subsequently conducted (round two). This phase intended to empirically investigate the impacts of environmental, economic, social, and administrative aspects of CDWM on WMH including reduce, reuse, recycle, and disposal strategies in the Malaysian context. The main questionnaire included two parts: personal information, and questions on the effectiveness assessment of the factors (Appendix A). Respondents were requested to rate their responses on the following scale: from 1 = not important at all to 5 = very important. The scaling technique for the main questionnaire was 5-point Likert scale since measuring subjective responses through this scales are extremely popular and easy to administer [98]. As discussed earlier, surveys have the advantage of being simple for designing, disseminating and analyzing data in an unbiased way [99]. Once the main questionnaire was designed, its content, clarity, efficiency in terms of time management, and quality were assessed by two academics and two professionals with extensive waste management experience and minor amendments were made according to their recommendations.

In this study, the target population was G7 construction contractors in Selangor, Malaysia, the highest grade of classification based on the Construction Industry Development Board (CIDB) registration for contractors involved in mega construction projects with an annual turnover of RM 200 million and above [100,101]. After evaluating the construction contractors with respect to the companies’ specific area of expertise (CDW) and relevant certificates, 132 companies left as the target population. Therefore, 132 questionnaires were sent to the selected companies through 2 ways, namely either on-site distribution and web-based survey in case it was impossible to perform on-site investigation, to ensure that the adequate response rate is achievable. Finally, 63 usable surveys were returned indicating an acceptable response rate of 47% [102].

Respondents were mostly male (87%) and between 38 to 46 years old (36.5%). Additionally, 59% of respondents reported having over 9 years of relevant experience in CDWM. By academic qualification, more than 80% of respondents had at least a bachelor’s degree. Considering respondents’ occupation, government officers were the most respondents with 33%, followed by construction research institutes of 27%. Respondents’ profile is summarized in

Table 3. Profile of respondents.

Gender	Female	Male
Number	8	55
Percentage (%)	13%	87%
Age	18–26	27–35	36–44	45–55	56 and above
Number	8	16	23	12	4
Percentage (%)	12.7%	25.4%	36.5%	19%	6.35%
Work Experience in CDWM	Less than 3	3–6	6–9	9–12	12 years and more
Year(s)	0	14	12	22	15
Percentage (%)	0	22%	19%	35%	24%
Highest qualification	Certificate	Diploma	Bachelor’s degree	Master’s degree	PhD
Number	3	9	20	17	14
Percentage (%)	4.8%	14.3%	31.7%	27%	22%
The most relevant role	Contractor	Academic	Consultant	Government officer	Construction research institutes
Number	10	6	9	21	17
Percentage (%)	16%	10%	14%	33%	27%

.

Structural equation modelling (SEM) is preferred over other data analytical techniques (i.e., factor analysis, regression analysis, etc.) since SEM can construct unobservable latent variables, deals well with the modelling errors, utilizes empirical data for testing a priori theoretical assumptions, and has an outstanding capability in measuring and assessing models among multiple variables [103]. Partial least squares-SEM and covariance-based-SEM (CB-SEM) are the two main classifications of SEM. Nevertheless, this study utilized PLS-SEM as it deals well with small sample sizes and complicated research models, handles missing data well, and has a potent statistical power [103,104]. Furthermore, as articulated by Hair Jr, Hult, Ringle and Sarstedt [103], PLS-SEM works well with less developed theories. SmartPLS version 3.0 software was utilized in this study for the purpose of data analysis.

4.3. Primary Data Tests

Collected data were primarily tested in opposition to three criteria of missing values, suspicious pattern in responses, and outliers [103]. Missing data refers to failing to answer a question either inadvertently or purposely [103]. Missing data should be identified and studied as missing data are nuisance to the results. It is generally recommended that missing completely at random as less than 10% of each individual item is acceptable and can be ignored. A frequency test of collected data showed that the missing data was under 1 percent and, in some items, no missing data was observed. Therefore, the dataset had no issue regarding the missing values or any biasness resulted from missing values. Also, no suspicious patterns of diagonal or straight lining were observed in the dataset [103]. In the last step, data were checked against outliers, which is referred to extreme response to a question in a completely different way than to other questions. In this regard, box plots and stem-and-leaf plots using IBM SPSS Statistics 23 software were considered. However, no notable outliers were observed [75,105]. It is also noteworthy considering that PLS-SEM handles non-normally distributed data because it employs bootstrapping, a non-parametric technique, that allows examining the statistical significance of different PLS-SEM results [103].

4.4. Assessing the Research Model

The research model encompassing measurement and structural model were assessed in this step. The measurement model was assessed prior to the structural model assessment, because unreliable or invalid constructs end up in unreliable or invalid relationships among constructs in the structural model [103]. Thus, reliability and validity of measures were assessed by performing internal consistency, convergent validity (AVE), and discriminant validity assessment [106,107]. The structural model was then evaluated considering five criteria of collinearity (VIF), coefficient of determination (R²), effect size (f²), path coefficients, and blindfolding and predictive relevance (Q²) [103,106].

5. Results

5.1. EAHP Results

Based on

Table 4. Global weight and ranking of each factor.

Criteria	Local Weight (1)	Code	Sub-Criteria	Local Weight (2)	Global Weight (3)	Ranking
1. Environmental aspect	0.29400	ENP1	1. Land consumption due to waste	0.05875	0.01897	20
		ENP2	2. Noise pollution	0.1632	0.04798	7
		ENP3	3. Water pollution	0.10354	0.03044	11
		ENP4	4. Air pollution	0.28028	0.08240	2
		ENP5	5. Public environment contamination due to illegal waste dumping	0.39424	0.11590	1
2. Economic aspect	0.32229	ECP1	1. Waste sorting, collection, and separation costs	0.18165	0.05866	3
		ECP2	2. Costs associated with reusing waste	0.1697	0.054806	5
		ECP3	3. Costs associated with recycling waste	0.1697	0.054806	5
		ECP4	4. Costs associated with waste transportation	0.08758	0.028285	13
		ECP5	5. Costs associated with waste disposal	0.06392	0.020643	16
		ECP6	6. Penalty for illegal dumping of waste	0.04288	0.013848	22
		ECP7	7. Revenue from selling recycled materials	0.17019	0.054964	4
		ECP8	8. Saving in project costs	0.07115	0.022978	15
		ECP9	9. Incentive mechanism for waste management	0.04322	0.013958	21
3. Social aspect	0.20143	SOP1	1. Waste management practitioners’ awareness	0.20089	0.03847	9
		SOP2	2. Provision of job opportunities	0.05144	0.00985	23
		SOP3	3. Arduous physical working condition	0.17887	0.034257	10
		SOP4	4. Health related long impacts	0.09788	0.018746	17
		SOP5	5. Waste management operatives’ safety	0.15165	0.029044	12
		SOP6	6. Public satisfaction about waste management practices	0.09622	0.018428	18
		SOP7	7. Public demand for regulating illegal waste dumping	0.09593	0.018372	19
		SOP8	8. Aesthetic impacts of illegal waste dumping	0.12712	0.02434	14
4. Administrative aspect	0.18228	ADP1	1. Policy quality	0.24771	0.04744	7
		ADP2	2. Technology	0.27371	0.05242	6
		ADP3	3. Management and supervision	0.20487	0.03923	8
		ADP4	4. Responsibility issues	0.27371	0.05242	6
Total	1.000		Total	1.000

, the main results of the EAHP survey could be summarized as the economic aspect of CDWM was seen the essential parameter in assessing sustainability aspect of CDWM by 0.32229 mean weight. The following criteria with 0.29400 mean weight was environmental aspect of CDWM. The third and fourth criteria were social and administrative aspects of CDWM with 0.20143, and 0.18228 mean weights, respectively. Also, ‘Provision of job opportunities’ and ‘Public environment contamination due to illegal waste dumping’ were the least and most important factors with 0.00985 and 0.11590 GW, respectively. After performing validity test of factors, seven experts advised to remove four factors with the lowest ranks. Therefore, one factor from environmental, two factors from economic, and one factor from social aspects, namely ‘Land consumption due to waste (ENP1)’, ‘Penalty for illegal dumping of waste (ECP6)’, ‘Incentive mechanism for waste management (ECP9)’, and ‘Provision of job opportunities (SOP2)’ were removed in this step. The remaining factors (22 factors) were entered into the second round (main questionnaire design and distribution), to assess and analyze their impacts on WMH in the Malaysian context.

5.2. Measurement Model Assessment

The measurement model assessment was performed to assess the measures’ reliability and validity [103]. The results of the measurement model evaluation using the PLS-SEM algorithm are represented in Figure 3.

The recommended threshold was set at x > 0.70 for outer loadings, x > 0.60 for Cronbach’s alpha (CA), x > 0.60 for composite reliability (CR), and x > 0.50 for average variance extracted (AVE) [106]. Loadings of more than 0.50 needed not to be eliminated if this elimination did not change the composite reliability (e.g., ADP4). Furthermore, items with low loadings were eliminated where their elimination improved the AVE [75,103]. Considering the results of the measurement model evaluation, SOP3 was the only indicator removed from further analysis. Results revealed that outer loadings, CA, CR, and AVE were all within the accepted threshold.

Table 5. Result of the measurement model.

Construct	Items	Loading	Cronbach Alpha (CA)	CR	AVE	N of Deleted Indicators
Environmental aspect	ENP2	0.849	0.786	0.858	0.605	0
	ENP3	0.678
	ENP4	0.856
	ENP5	0.711
Economic aspect	ECP1	0.814	0.891	0.915	0.606	0
	ECP2	0.832
	ECP3	0.723
	ECP4	0.852
	ECP5	0.716
	ECP7	0.799
	ECP8	0.697
Social aspect	SOP1	0.769	0.793	0.854	0.543	1 (SOP3)
	SOP4	0.621
	SOP5	0.624
	SOP6	0.825
	SOP7	0.816
	SOP8	0.734
Administrative aspect	ADP1	0.896	0.787	0.860	0.610	0
	ADP2	0.824
	ADP3	0.750
	ADP4	0.629
WMH	Reduce	0.709	0.605	0.772	0.500	0
	Reuse	0.738
	Recycle	0.723
	Disposal	0.527

represents the results for the measurement model evaluation.

Discriminant validity was evaluated in the last step of measurement model assessment. As recommended by Hair Jr, Hult, Ringle and Sarstedt [103], heterotrait–monotrait ratio (HTMT) is the most vigorous and effective test for discriminant validity assessment, which was conventionally assessed through Cross-Loadings and Fornell-Larcker criterion tests. Complete bootstrapping with 5000 iterations was used to determine the HTMT values [106]. The results for HTMT are represented in

Table 6. Discriminant validity of SCDWM-HTMT.

HTMT (Bootstrapping)
Constructs	Administrative	Economic	Environmental	Social
Economic	0.423
Environmental	0.527	0.498
Social	0.658	0.545	0.608
WMH	0.789	0.515	0.664	0.696

. As HTMT values were below 0.85, the results of discriminant validity were accepted [106].

5.3. Structural Model Assessment

The first step towards the structural model assessment is to ensure that the items do not have linear relationship [108]. Based on Hair Joseph, Risher Jeffrey, Sarstedt and Ringle Christian [106], variation inflation factor (VIF) values were all acceptable as they were below 5. In the second step, in order to assess the predictability power of the research model, included and excluded coefficients of determination (R²) were evaluated. Since the R square values were above 0.5, the predictability power of the research model was revealed substantial [106].

In the third step, the effect size (f²) of constructs were calculated. Effect size (f²) shows the relative contribution of each exogenous variable to the R square of the target construct. According to Cohen’s table of effect size [109], any value greater than 0.02 demonstrate a small effect size, greater than 0.15 a medium effect size, and greater than 0.35 demonstrate a large effect size. Thus, the economic aspect of CDWM with 0.637 indicated a large effect size, whereas social (0.289) and environmental (0.222) aspects of CDWM represented medium effect size. Finally, the administrative aspect of CDWM displayed a small effect size (0.052).

In the fourth step, bootstrapping technique with 5000 iterations for two-tailed test with a significance of 0.1 was applied to attain structural path coefficients (p or t values), in which t-values of 1.65 and more was set as the basis of hypotheses acceptance [106]. It was revealed that the economic aspect of CDWM (t-value: 6.833) had the greatest impact on WMH. Thus, hypothesis H₁ was supported. Similarly, social (t-value: 4.678), environmental (t-value: 4.142), and administrative (t-value: 1.831) aspects of CDWM were proven effective on WMH, respectively.

In the final step, Stone–Geisser’s (Q²), which defines as the predominant measure of predictive relevance that postulates a model to properly predict each endogenous latent construct’s indices was calculated. Thus, blindfolding process in SmartPLS was run. Since, the sample size was N = 63, a D value of 4 was chosen since the remainder of the ‘63’ divide by ‘4’ is different from zero. A blindfolding process applies only to endogenous latent constructs [75,103,108].

Table 7. Collinearity assessment, effect size, path coefficients, and predictive relevance.

1. Collinearity Assessment (VIF), 2. Coefficient of Determination (R2), and 3. Effect Size $\left({\mathit{f}}^2\right)$
Indicators/Variables	VIF	Rin²	Rex²	${\mathit{f}}^2$	Effect Size
Environmental aspect		0.865	0.835	0.222	Medium
ENP2	1.330
ENP3	1.510
ENP4	1.498
ENP5	1.858
Economic aspect		0.865	0.779	0.637	Large
ECP1	2.210
ECP2	2.076
ECP3	1.640
ECP4	2.316
ECP5	1.708
ECP7	2.028
ECP8	2.355
Social aspect		0.865	0.826	0.289	Medium
SOP1	1.752
SOP4	1.438
SOP5	1.407
SOP6	1.847
SOP7	1.560
SOP8	1.346
Administrative aspect		0.865	0.858	0.052	Small
ADP1	1.539
ADP2	1.428
ADP3	1.424
ADP4	1.230
WMH		-	-	-	-
Reduce	1.158
Reuse	1.097
Recycle	1.111
Disposal	1.014
4. Path Coefficients
Hypotheses	Relationship			t-Value	Supported
H1	Environmental → WMH			4.142	Yes
H2	Economic → WMH			6.833	Yes
H3	Social →WMH			4.678	Yes
H4	Administrative → WMH			1.831	Yes
5. Predictive Relevance
WMH			0.353

summarizes the results for collinearity assessment, effect size (f²), hypotheses testing, and predictive relevance (Q²) assessment.

6. Discussions

This paper set out to investigate and provide a better concept of SCDWM in the Malaysian construction industry by exploring the different indicators of each contributing factor to SCDWM and to distinguish whether these factors could potentially affect WMH.

6.1. Synthesis of Empirical Findings

This study empirically analysed SCDWM in the Malaysian context. In the first step, by meticulously reviewing the relevant literature to CDWM, 26 factors under categories of environmental, economic, social, and administrative aspects of CDWM were identified as being effective in managing CDW. Then, the EAHP method was applied to prioritise and refine these factors. In this step, from 26 primarily identified factors, 4 factors including ENP1, ECP6, ECP9, and SOP2 (Table 3) were removed based on their low ranks, followed by experts’ recommendations.

EAHP results indicated that economic aspect of CDWM was the most important criterion considering its highest mean weight, followed by the environmental aspect of CDWM. Also, the social and administrative aspects of CDWM were the third and fourth criteria, respectively. The order of economic and environmental aspects of CDWM was consistent with a study performed by Wu, et al. [110]. Furthermore, it was revealed that ‘Provision of job opportunities’ and ‘Public environment contamination due to illegal waste dumping’ were the least and the most important factors, respectively. Also, ‘Waste sorting, collection, and separation costs’, ‘Waste management practitioners’ awareness’, and ‘Technology’ along with ‘Responsibility issues’ were the most effective factors of economic, social, and administrative aspects of CDWM, respectively, which was consistent with study performed by Henri and Journeault [111] in terms of significance.

Afterwards, 22 factors involved in CDWM from the sustainability perspective along with 4 factors reflecting WMH (reduce, reuse, recycle, and disposal strategies), were entered into the main questionnaire design and distribution for the purposes of assessing their impacts to decide on whether accept or reject formulated hypotheses. With the help of SmartPLS software, data were analysed, and the measurement model and structural model were assessed. Factor SOP3 was removed in the process of measurement model assessment as it was below the accepted threshold of 0.5.

Results obtained from bootstrapping for path coefficients revealed that the impact of the economic aspect of CDWM on WMH is significant (largest effect), as its t-value (6.833). This finding is consistent with the results of a study performed by Wu, Yu and Shen [110]. Other scholars have also emphasized the pivotal role of economic factors, such as costs associated with waste collection, transportation, recycling, disposal and illegal dumping on waste management strategy (e.g., recycle strategy) selection [34,55,69,112].

Results also indicated that the social aspect of CDWM had the second largest effect on WMH (t-value: 4.678). The significance of social aspects of sustainability in CDWM has been proven by many scholars [70,113]. Results obtained from data analysis showed the significance of the social aspect of sustainability in waste management. This result is aligned with studies in which the importance of factors such as waste management stakeholders’ health and safety, public health and safety, collaboration and communication among waste management professionals and public, etc. have been proven effective in waste management [5,70,81]

Accordingly, the environmental aspect of CDWM had the third largest effect among all other factors (t-value: 4.142). Although this impact is not as strong as economic and social impacts, it has a medium to large effect on the overall model, which is consistent with the findings of [44]. Factors such as environmental pollution, environmental impacts of landfilled waste, and detrimental impacts of illegal waste dumping have been proven effective in WMH in many studies [10,114].

In the last step of hypotheses testing, the administrative aspect of CDWM was found to be the least influential factor in determining the variation in WMH (t-value: 1.831). Although the role of the administrative aspect of CDWM in SCDWM with respect to its application in this study has not been applied in similar studies, if it corresponds to the technical and technological issues, waste management supervision, responsibility and management issues of stakeholders, and CDWM regulations, it is consistent with previous studies [10,21]. In this regard, although CDWM regulations such as national strategic plan for solid waste management, national recycling program, waste recycling incentives [115], recycling technologies [116], etc. have already been set up in Malaysia, SCDWM calls for advanced application and coordination in these areas.

6.2. Theoretical Contribution

This study, initially, developed a model considering the relationships between contributing factors to sustainability aspect of CDWM and WMH. In this study, CDWM was meticulously revisited to understand further the dynamic nature of SCDWM that could potentially lead to a more SCDWM. Theoretically, the developed conceptual framework in this study is based on the integration of sustainable development and waste management, which has resulted in the emergence of sustainable waste management (SWM), which was later evolved and formed the waste management hierarchy (WMH), the most effective strategy for managing CDW. This model significantly contributed to a conception of the determinant factors of sustainability within CDWM and WMH. Furthermore, this study contributed to CDWM by introducing economic, environmental, social, and administrative aspects of CDWM as the pillars of SCDWM and determinants of WMH. Finally, the findings of this paper added to theoretical contributions by proposing and validating a model that could predict WMH. Results indicated that economic, social, environmental, and administrative aspects of CDWM effectively affect WMH, respectively, however, the role of the administrative aspect of CDWM should receive more emphasis in future studies.

6.3. Practical Contribution

The findings provided a basic guideline for facilitating the process of migration from contributing factors to the sustainability aspect of CDWM to WMH. The findings indicated that all factors contributed significantly to WMH. The economic aspect was the most influential factor in managing CDW sustainably in the Malaysian construction industry, which simply means that the policy and decision makers should care to obviate the economic costs and increase the revenue from CDW. In other words, if decision makers are looking for a swift and simple policy, the answer is to provide economic incentives for the CDW practitioners. Environmental and social aspects of CDWM also contributed significantly to WMH. Although, these effects were not as strong as the economic aspect, they have moderately or highly affected the overall model. Therefore, a good policy and long-term plan needs to take the environmental and social impacts into account as much as economic impact. Similarly, much attention needs to be given to administrative factor to find the big gaps and address challenges associated with CDW regulations, supervision, and management. In fact, the model developed in this study could be applied as a self-assessment tool by the construction industry parties to better understand their positions and levels of sustainable practices in managing CDW. The determinants of SCDWM are beneficial for authorities to improve their strategy, sustainability policy, and practice for metering the rising requirements for sustainable development in construction industry.

7. Conclusions

This study provided a thorough comprehension of SCDWM, particularly in the Malaysian construction industry. Upon meticulously reviewing literature pertinent to SCDWM, 26 factors under 4 categories of economic, environmental, social, and administrative aspects of sustainability with respect to CDWM were initially identified. The factors underwent pilot testing by the application of EAHP. In this step, the refining process of factors was carried out based on the factors’ weights and experts’ judgements. In this step, four factors including ENP1, ECP6, ECP9, and SOP2 (Table 3) were removed from further analysis. In the next step, in order to empirically investigate the impact of sustainability aspects of CDWM on the most effective strategy for waste management (also known as WMH and including reduce, reuse, recycle and dispose strategies), the main survey was designed and distributed to 132 G7 construction companies in Malaysia with the most relevant experience and expertise in CDWM, to collect data based on a five-point Likert scale. This process was undertaken to analyse the research model including four hypotheses that reflected the impacts of the environmental, economic, social, and administrative aspects of CDWM on WMH.

After performing data analysis through PLS-SEM by SmartPLS version 3, it was concluded that economic factor is the most influential factor (t-value: 6.833) in determining WMH, followed by the social (t-value: 4.678) and environmental (t-value: 4.142) aspects of CDWM, respectively. Also, it was revealed that the administrative aspect of CDWM had the least effect in determining WMH (t-value: 1.831). Therefore, it was suggested that we look at the determinants and mediating effect of this factor on SCDWM. In line with these findings, it was concluded that a swift and simple policy could focus on the economic aspect of CDWM, since it was the most effective aspect of SCDWM, while a holistic policy and a long-term plan needs to consider the most influential factors as economic, environmental, social, and administrative factors.

Future studies can be built on this model and develop new factors or moderators, (e.g., a different policy to provide better understating of challenges and contributing factors to SCDWM), which represents a limitation of current study. In terms of administrative factors, this paper has found a small effect for this factor, proposing further research on developing more related items for this construct, as well as identifying more significant contributors to SCDWM.