Green construction practices: ensuring client satisfaction through health and safety performance

Studies have proven that client satisfaction (CS) is greatly affected by the nature of construction practices on site. However, not much attention has been given to CS in projects that adopt green construction practices (GCPs). This study examines the effect of GCPs on CS, and the means by which CS can be achieved through health and safety performance (HSP). This study also analyses how the relationship between HSP and CS could vary depending on the level of GCPs adopted. A questionnaire with 27 items drawn from literature was used in collecting data from class A contractors in Nigeria. The study’s hypotheses were tested using the partial least squares structural equation modelling (PLS-SEM) technique. The findings of the study reveal that GCPs have a positive effect on CS. HSP proved to only partially mediate the relationship between GCPs and CS. The findings also show that the effect of HSP on CS is moderated by GCPs with the stronger effect from projects with low adoption of GCPs. The study’s findings highlight the pivotal role of HSP in ensuring CS in construction projects that adopt GCPs. This study makes practical contributions to the body of existing knowledge by highlighting the need for concerted efforts by contractors such as investing in health and safety training and ensuring site workers wear personal protective equipment aimed at adhering to acceptable health and safety standards. This can be achieved through provision of funds for safety activities, monitoring, and supervision of workers to guarantee HSP which translates to CS while adopting GCPs. However, the obstacles to these are the lack of safety culture by construction organizations, lack of enlightenment on the part of construction site workers, and poor safety climate on the construction project site. Additionally, the study provides greater insight on the practice of green construction among contractors by showing that the effect of HSP on CS would differ considerably on account of the level of GCPs adopted.


Introduction
The construction sector has been one of the main contributors to the environmental problems been witnessed today due to the adoption of traditional construction methods. These environmental problems have been on the upward trend due to rapid urbanization resulting from rural-urban migration (Ekins and Zenghelis 2021;Meng et al. 2021). However, the construction sector contributes enormously to the economy of any country. In order to solve the environmental issues arising from construction activities, the adoption of green construction practices (GCPs) has become imperative (Lu et al. 2021;Tavana et al. 2021). In fact, GCPs have received special attention from researchers, contractors, and governments because it is one of the most important and effective means of reducing environmental problems (Badi and Murtagh 2019).
Due to the importance of GCPs in reducing environmental problems and ensuring environmental sustainability, various environmental laws and rating systems have been developed in different countries. In the USA, the Leadership in Energy and Environmental Design (LEED) was developed, the UK have a green rating tool (Atanda and Olukoya 2019). However, due to the need for environmental protection, the Federal Environment Protection Agency (FEPA) was created by the federal government of Nigeria in 1988. FEPA is saddled with the responsibility of providing guidelines and minimum standards for environmental protection (FEPA 1991). The guidelines and standards provided by FEPA cover issues such as waste management, water quality, management of harmful substances, and air quality. Subsequently, additional environmental regulations and standards were put in place with the primary aim of environmental protection. These include the National Policy on the Environment in 1989, the National Environmental Standards and Regulations Enforcement Agency Act 2007, and the National Building Energy Efficiency Code of 2017.
Though these environmental laws and regulations were designed to ensure that GCPs are adopted, the satisfaction of the client is often neglected (Bai and Satir 2020). It should be noted that although the emphasis of GCPs basically considers the environmental impact of construction activities, the social aspect of sustainability in which client satisfaction (CS) is a vital part must not be ignored (Murtagh et al. 2020). Hence, the importance of CS in the adoption of GCPs was emphasized by Athapaththu and Karunasena (2018) and Olawumi and Chan (2021). Athapaththu and Karunasena (2018) stated that contactors rarely opt to adopt GCPs except if the client insists it should be adopted. Therefore, the crucial role played by CS in GCP adoption is not in doubt. In view of the above, Shurrab et al. (2019) advocated that investigation ought to be conducted to determine the impact of GCPs on CS.
GCPs may lead to negative, positive, or no effect on CS. However, the strength, direction, and nature of this relationship need to be examined empirically to gain greater insight on the relationship between GCPs and CS, which could be of benefit to contractors engaged in projects that adopt GCPs (Alam and Islam 2021;Baah et al. 2021;Shurrab et al. 2019;Suganthi 2019). Also, studies such as Cianciarullo (2019), Onubi et al. (2021), Zhou et al. (2015), and Khan et al. (2021) have pointed to the possible positive effect of GCP on the health and safety performance (HSP) of construction projects, while El-Sayegh et al. (2019), Abu Aisheh et al. (2021), and Karakhan and Gambatese (2017) observed a positive effect of HSP on CS. Nevertheless, there is a need to comprehensively analyse the possibility of the mediating effect of HSP on the relationship between GCPs and CS, with a view to understanding the path through which CS can be achieved while adopting GCPs. In other words, the mechanism through which GCPs could lead to CS needs more detailed evaluation. Thus, this study introduces HSP in the analysis of the GCPs-CS relationship. Furthermore, studies have revealed that the strength and direction of the effect of HSP on CS could differ depending on the level of GCPs adopted (Wuni et al. 2019), but more empirical details are needed to further validate and verify this claim with a view to providing a clear perspective on how this occurs on construction projects (Liang and Shi 2021). Hence, in order to contribute theoretically to existing literature, this study considered GCPs as a moderator in the relationship between HSP and CS. Overall, this research is aimed at examining the effect of GCPs on CS, the mediating effect of HSP on the relationship between GCPs and CS, and the moderating effect of GCPs on the HSP-CS relationship.
The contributions and significance of this paper are threefold. Firstly, it presents the direct and mediating (indirect) effects of how GCPs and HSP influence CS, thereby allowing us to have an all-encompassing view of the intricacies that guarantee CS on projects that adopt GCPs. Secondly, the study explores the moderating effect of GCPs on the relationship between HSP and CS, thus giving contractors and academic scholars a holistic view of how CS resulting from HSP could differ depending on the level of GCPs adopted. Lastly, the results of this study will trigger the formulation of effective policies and programs by the government to encourage contractors to adhere to HSP guidelines while adopting GCPs for optimal CS. Precisely, the findings of this paper will be of benefit to contractors, clients, and other stakeholders involved in the adoption of GCPs on construction project sites.

Concept of green construction practices
The terms "green construction", "cleaner construction", and "sustainable construction" are frequently used interchangeably (Shurrab et al. 2019). GCPs can be defined as practices undertaken with the intention of ensuring construction quality and safety, conserve resources, and reduce the harmful effect of construction activities on the environment through energy management, land management, water savings, and materials management (Iqbal et al. 2021;Jiang et al. 2019;Rahman et al. 2021). Additionally, Jiang et al. (2019) defined GCPs as an all-encompassing view of sustainable development in the construction industry. Similarly, Shurrab et al. (2019) refer to GCPs as the use of onsite practices aimed at minimizing the negative environmental impact of construction activities. Other authors such as Wang et al. (2018) and Alabi et al. (2021) described cleaner construction as construction practices that are focused on the reduction of the adverse impact of construction activities on the environment, reduction in the use of non-renewable resources, and lesser energy consumption during construction processes. While sustainable construction can be said to be a broad concept encompassing environmental, economic, and social aspects of sustainability in the construction industry (Kiani Mavi et al. 2021). Also, Ogunbiyi et al. (2014, p. 84) stated that "sustainable construction refers to the set of processes by which a profitable and competitive industry delivers built assets (buildings, structures, supporting infrastructure, and their immediate surroundings), which: enhance the quality of life and offer customer satisfaction; offer flexibility and has the potential to cater for user changes in the future; provide and support desirable natural and social environments, and maximize the efficient use of resources". It should be noted that green construction and cleaner construction have the same meaning (Pham et al. 2021). They both refer to the environmental aspect of sustainable construction. In other words, green construction and cleaner construction are subsets of sustainable construction.
The idea behind GCPs transcends the construction of buildings without posing any negative effect on the natural environment since it seeks to create conditions that ultimately lead to the improvement in the quality of the environment . Simply put, GCPs are a combination of attempts geared towards building in a responsible manner while reducing waste, controlling material usage, managing energy use, managing stormwater, and preserving the environment. GCPs take into consideration conservation needs and utilizes resources more efficiently than conventional construction practices (Ofek et al. 2018).

Green construction practices and client satisfaction
CS is a measure of the accomplishment of the expectations the client had before the start of the project, and the actual performance which can be measured at the various phases of the construction project (Li et al. 2013). CS has gained popularity in the measurement of project performance as an addition to the traditional performance criteria of cost, quality, and time (Davis 2016).
CS is of utmost importance in the adoption of GCPs (Presley and Meade 2010). However, many clients today do not fully appreciate the benefits of GCPs due to their seeming lack of understanding of its merits (Shan et al. 2020). This has led to a situation whereby clients become doubtful about the capacity of GCPs to satisfy their needs . This is in tandem with the findings of Windapo (2014) who opined that clients are not likely to adopt GCPs for the purpose of environmental protection without government pressure and any financial benefits accruable to them. Even though the number of clients accepting to adopt GCPs has been on the increase, their choice to adopt these practices is passive, since they have been compelled by the government to do so (Shan et al. 2020). This could lead to a situation whereby they may not be satisfied as they ought to be when such practices are adopted on their projects due to cost overrun, delays, increase in complexity, and risks (Onubi et al. 2019).
Adopting GCPs has been reported to be associated with an increase in the scope, increase in complexity, and the use of novel technologies in construction projects which may increase project costs and decrease the anticipated profit of the client (Ofek et al. 2018), thus negatively impacting on CS. Consequently, there has been a seeming lack of interest by clients in GCPs (Venkataraman and Cheng 2018). Additionally, most clients intend saving money in every project regardless of whether GCPs are adopted or not; however, most of the green construction projects' cost surpasses the original contract sum (El-Sayegh et al. 2019;Zhao et al. 2020), hence creating an inverse relationship between GCPs and CS. A similar study conducted in the tourism sector on the effect of green practices on CS reported that there exists no significant effect (Assaker 2020).
Lack of efficiency in the adoption of GCPs and nonparticipation of clients in the green construction process could be responsible for their lack of satisfaction (Li et al. 2018). Additionally, Moise and Gil-Saura (2020) and Issock et al. (2019) state that clients' perception of value is a precursor to their satisfaction. Clients are mindful of money spent on green products, and if they feel that adopting these green practices is not value for money, they will be dissatisfied (Moise and Gil-Saura 2020). Issues associated with green construction such as higher design cost and inability to meet projected sustainability targets can have negative effects on CS (Ayman et al. 2020). Gallarza et al. (2016) stated that the clients' value for money is positively associated with cognitive satisfaction. Therefore, when a client perceives that engaging in GCPs does not give value for money, they will not be satisfied with its outcome. The lack of satisfactory management system in green construction process, which entails revealing green product information to clients, and communicating effectively and consistently with clients during the construction process, has proven to be responsible for the negative relationship between GCPs and CS (Jiang and Wong 2016).
Additionally, the importance of schedule performance in any construction project and the role it plays in CS cannot be overstated. Gurgun and Koc (2020) report that compliance to schedule requirements was topmost among clients' performance criteria in GCPs. However, it has been proven that projects adopting GCPs are often more susceptible to schedule overruns compared to conventional construction projects (Bakchan et al. 2019), thus impacting negatively on CS on green construction projects. In view of the leanings from the literature reviewed, it is hypothesized as follows: H1: GCPs have a negative significant effect on CS.
Mediating effect of health and safety performance on the relationship between green construction practices and client satisfaction A construction project's HSP is a measure of the level to which both the primary and secondary stakeholders of the project are exposed to or affected by accidents, get injured, and/or are susceptible to other health-related issues resulting from construction activities (Ammad et al. 2020). The HSP of construction projects is considered a vital part of green construction (Hinze et al. 2013). However, early studies related to safety in green construction have reported that the adoption of GCPs poses greater safety risks to construction workers compared to conventional construction methods (Hwang et al. 2018). Also, Dewlaney and Hallowell (2012) found that the unfamiliar tasks, new materials, innovative technologies, and processes involved in GCPs are responsible for the safety risks reported in previous studies. Similarly, Karakhan and Gambatese (2017) observed that the severity of safety risk and exposure to hazards of construction workers is heightened when GCPs are adopted.
Contrary to the above, Cianciarullo (2019) opined that GCPs can lead to HSP with adequate planning and experienced construction workers. For example, Zhou et al. (2015) opined that HSP can be achieved while adopting GCPs if necessary measures are taken at the design stage of the projects. Similarly, Karakhan and Gambatese (2017) opined that the safety issues associated with GCPs can be prevented through strategies adopted during design. As such, many green building rating tools such as LEED have included credits for the prevention of safety issues through design (Karakhan and Gambatese 2017). In view of the increased quest and measures taken towards HSP in projects that adopt GCPs, Khan et al. (2021) report that there has been a significant improvement in the HSP during GCPs adoption. Yusliza et al. (2020) report that GCPs are largely expected to lead to HSP. Also, Sang et al. (2018) acknowledged that the technology involved in GCPs is complex but not highly unsafe.
Specialization in job functions reduces the frequency of accidents (Stoilkovska et al. 2015). It has been reported that if workers specialize and continue to adopt certain practices on the construction site over time, they become used to the rudiments involved leading to better HSP outcomes (Enshassi et al. 2019). Hence, since workers involved in the adoption of GCPs are expected to have been engaged in these practices over time, the likelihood of health and safety issues occurring is minimized (Man et al. 2019). The level of HSP recorded on clients' projects would positively influence their image in the society (Staniškienė and Stankevičiūtė 2018). Therefore, with guaranteed HSP resulting from GCPs, the client's reputation and image will be improved, thus leading to CS.
The HSP benefits resulting from the adoption of GCPs are anticipated to yield both monetary and non-monetary benefits such as cost benefits, schedule performance, increased project quality, and greater workers' productivity (Howarth and Watson 2010). Due to the merits of completing the project within schedule, without cost overruns, without accidents, and complying with government environmental regulations, the clients are expected to be satisfied with the project delivery process (El-Sayegh et al. 2019). Also, since GCP techniques and methods have been considered to be more expensive, more time-consuming, and more complex than the traditional construction methods thus leading to client dissatisfaction and reluctance to adopt these practices (Zhao et al. 2020), it is anticipated that the guaranteed HSP emanating from the adoption of GCPs as stated earlier would reduce the cost burden originating from safety issues, risks, less delay due to fewer accidents, and less absenteeism thus leading to CS. In other words, GCPs can lead to CS through HSP. Consequently, we hypothesize as follows: H2: HSP mediates the relationship between GCPs and CS.
Moderating effect of green construction practices on the relationship between health and safety performance and client satisfaction In the preceding section, previous studies had shown that HSP has a positive effect on CS. However, we contend that the nature and the magnitude of the effect will differ considerably depending on the level to which GCPs are practiced on the construction projects. In a study carried out in Nigeria, Oladokun et al. (2020) found out that construction professionals view GCPs adoption as the major determinant for CS in the country. Client demand has been tagged as the major driver that propels contractors to adopt GCPs (Athapaththu and Karunasena 2018). Similarly, Shurrab et al. (2019) opined that clients are in the know of the merits of GCPs as a means of preserving the environment and are willing to pay the required cost. Ochieng et al. (2014) asserts that CS is a measure of sustainability on construction projects as clients are increasingly demanding that contractors adopt GCPs. All these show that construction clients acknowledge the importance of adopting GCPs and demand for it. Nonetheless, there is a need to examine the moderating effect of GCPs on the relationship between HSP and CS in the construction field with a view to providing contractors with details on how the HSP-CS relationship could vary depending on the extent to which GCPs are adopted. In the tourism sector, Lee et al. (2018) reported a negative significant moderating effect of green practices on the relationship between service quality and CS. However, from the insights gotten from past studies, the positive effect of HSP on CS will be stronger with a higher GCPs adoption relative to projects with a low level of adoption of GCPs. Building on these arguments, it is hypothesized that: H3: The effect of HSP on CS will be stronger when GCP adoption is high.
The conceptual model of the study containing the study's hypotheses is shown in Fig. 1.

Research survey
To determine the effect of GCPs on CS, and the mediating role of HSP on the relationship between GCPs and CS, a survey of 206 construction projects of class A contractors completed between 30th of March 2015 and 28th of February 2019 was conducted. One questionnaire was administered per project with 168 useable responses received, accounting for an 81.55% response rate. The 168 useable responses received are adequate according to the minimum sample size requirement of 160 samples using the inverse square root method recommended by Kock and Hadaya (2018). The minimum sample size using the inverse square root method recommended by Kock and Hadaya (2018) was determined using Equation 1: where Ň = minimum sample size required and |β| min = absolute value of minimum path coefficient of 0.197. The solution to this equation yielded a minimum sample size of 160.
Class A contractors in Nigeria are determined by their registration status with the Bureau of Public Procurement (BPP). They are the highest class of contractors in Nigeria. The list of class A contractors was obtained from the BPP. Class A contractors unlike class B, C, and D are permitted by law to bid and execute construction projects whose monetary value exceeds ten billion naira ($27,777,777). In other words, class A contractors have no monetary limit to the cost of projects they are permitted by law to undertake. Contractors registered under class A are expected to have completed a minimum of twelve construction projects within 5 years before their registration with the BPP. Other requirements for contractors to be registered as class A include having a minimum of three mechanical engineers, one accountant, six land surveyors, six construction managers, and ten civil engineers all having at least 5 years of working experience.
The choice of class A contractors instead of other classes of contractors for this study is because previous studies such as Waris et al. (2014) and Chang et al. (2018) have shown that there exists a link between firm size and the adoption of GCPs, with large construction firms adopting GCPs more than smaller firms. They (class A contractors) are usually involved in big and complex projects: as such their construction activities may have more negative impact on the environment and CS than other classes of contractors. Also, the reason for the choice of "class A" rather than the entire contractors is because they are most likely to have verifiable addresses unlike lower level contractors (Yusof et al. 2020). However, there is no evidence to prove that the findings of this present study will systematically differ when other classes of contractors that adopt GCPs are considered.
The survey method was adopted in the data collection, and only projects executed by contractors graded as "class A" were considered using the convenience sampling technique, and all selected projects had adopted GCPs.

Research questionnaire and measures
The questionnaire has four parts. The first part has 5 items focusing on the respondents' profile. The second part had 12 items focusing on GCPs adopted on the construction site adapted from Ajayi et al. (2016), Collins et al. (2017), and Eaton (2018). The third part contained 5 items concerning HSP adapted from Kwon (2013), while the fourth section had 5 items related to CS adapted from Berssaneti and Carvalho (2015). In the second and third sections of the questionnaire, the respondents were asked to indicate their responses on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree), while the fourth section's responses were collected using a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). Also, an additional question termed "global item" representing a general response to all the items in the  constructs was asked per construct as recommended by Hair et al. (2017) for the purpose of redundancy evaluation. The adoption of different Likert scales (5 and 7 points) was to minimize the likelihood of common source bias (CSB) occurring using procedural remedies during data collection as recommended by Podsakoff et al. (2003). Details of the questionnaire items and their respective sources are contained in Table 1.
Since the data for all the study's constructs were collected from the same respondents, there is a possibility of CSB in the study which could result in misleading findings (Craighead et al. 2011). Therefore, aside from the procedural methods adopted previously, two statistical methods were adopted to examine CSB. First, the full collinearity variance inflation factor (FCVIF) recommended by Kock (2015) was used, and the result showed that the FCVIF value of 1.325 was obtained which is below the 3.3 threshold value recommended by Kock (2017). Secondly, the Harman single factor test (Harman 1960) was conducted and yielded a percentage variance of 5.492% which is substantially lower than the 50% cut-off value recommended by Harman. Thus, CSB is not an issue in this study.

Data analysis and results
The partial least squares structural equation modelling technique (PLS-SEM) was adopted in the data analysis for this study using the warpPLS 7 software. The PLS-SEM technique was used because the study involved theoretical framework testing from a predictive standpoint, and the data for this study were not normally distributed as evidenced from the results of Shapiro-Wilk test which was found to be significant at 0.000 (Hair et al. 2019). The moderation effect was analysed using the two-stage approach since the model for this study contains constructs that are all measured as formative and has a moderator (Fassott et al. 2016). For the analysis of the mediator, the transmittal approach recommended by Rungtusanatham Table 1 Questionnaire items and source

Constructs Source
Green construction practices (GCPs) Waste management

Stormwater management
Permeable pavement (GCPs9) Eaton (2018) Grass swales were used for stormwater management (GCPs10) Eaton (2018) The natural topography of the construction site was preserved (GCPs11) Eaton (2018) The practice of harvesting rainwater was adopted on the site (GCPs12) Eaton (2018)

Health and safety performance (HSP)
Hazard prevention measures were implemented on the project (HSP1) Kwon (2013) Safety performance was in line with the experience level of the workers (HSP2) Kwon (2013) Safety of the public was not compromised during construction works (HSP3) Kwon (2013) Safety performance was in line with anticipated safety targets (HSP4) All relevant safety guidelines were adhered to (HSP5) Kwon (2013) Client satisfaction (CS) Client(s) was satisfied with the project (CS1) Berssaneti and Carvalho (2015) The client was satisfied with the level of compliance with technical specifications (CS2) Berssaneti and Carvalho (2015) No rework was needed on the project (CS3) Berssaneti and Carvalho (2015) Failure of fixtures and fittings recorded was minimal (CS4) Berssaneti and Carvalho (2015) Clients were satisfied with the standards achieved (CS5) et al. (2014) was adopted. Hence, only the indirect effect is considered in the mediation analysis as suggested by Hayes and Rockwood (2017).

Demographic profile of respondents
The results obtained for the demographic profile of the respondent indicate that most of the respondents (29.2%) had a masters' degree, while those with a doctorate degree (Ph.D.) were the least making up 5.4% of the respondents. For the years of working experience, most respondents had between 11 and 15 years of working experience accounting for 28.6% of the study sample. Additionally, most of the respondents were site managers and made up 48.2% of the study sample. Regarding the number of full-time employees in the contractors' organization, most of the contractors had over 200 full-time workers accounting for 36.3% of the respondents.

Measurement model
In the analysis of the formative measurement model, the following were assessed: convergent validity, assessment of the model for collinearity issues, and significance and relevance of formative indicators. Table 2 shows that the convergent validity of all three constructs are greater than 0.7, which indicates that the level of convergent validity in the formative constructs is adequate (Hair et al. 2017). Regarding the multicollinearity between the indicators in the model, the values of the variance inflation factor (VIF) for all the indicators as shown in Table 2 are less than the 3.3 cut-off values suggested by Kock (2014) which signifies that collinearity in the model is within acceptable limits. For the significance and relevance of indicator weights, the results show that all indicators have significant weights apart from GCP4, GCP7, GCP9, GCP11, and HS4. However, the results presented in Table 2 shows that all the nonsignificant weights have indicator loadings that are greater than 0.5; therefore, the indicators are retained in the formative construct (Hair et al. 2017). After the analysis of the formative constructs, it is concluded that the model satisfies all the requirements for a formative measurement model.

Structural model
Prior to evaluating the structural model, the VIF was used to test for lateral collinearity among all the constructs. The VIF value for all three constructs ranges from 1.208 to 1.671, thus satisfying Diamantopoulos and Siguaw (2006) 3.3 threshold value. Hence, lateral collinearity is not a concern in the study (Table 3) (Hair et al. 2017).
The structural model is presented in Fig. 2 and contains the coefficient of determination (R 2 ) and the path coefficients (β). An R 2 value of 0.45 was obtained in this study indicating that the model explains 45% of the variances in GCPs and HS. Also, the R 2 value of 0.45 is more than 0.26 recommended by Cohen (1988) which signifies a substantial model. Additionally, the predictive relevance, Q 2 , values of 0.032 and 0.447 obtained are substantially greater than zero, signifying that the model has adequate predictive relevance (Hair et al. 2017).

Hypothesis testing for direct effect
The results of the hypothesis testing for the study's relationships are presented in Fig. 2 and Table 4. The result for the direct effect of GCP on CS is positive and statistically significant (β=0.347, P<0.001) but does not support the first hypothesis (H1) since the study hypothesized a negative significant effect. The effect size (f 2 ) of 0.15 obtained for this relationship is indicative of a medium effect going by Cohen (1988) guideline.

Hypothesis testing for mediating effect
For the mediating effect, the results presented in Table 4 indicate that the mediating relationship is significant (β=0.141, P=0.031) with a small f 2 of 0.032, t h u s s u p p o r t i n g t h e s e c o n d h y p o t h e s i s ( H 2 ) . Additionally, the type of mediation present in this study is partial mediation since both the direct (GCPs ➔CS) and indirect (GCPs ➔HSP➔CS) relationships are significant (Rucker et al. 2011). This implies that HSP has absorbed only a fraction of the direct effect of the independent variable (GCPs) on the dependent variable (CS).

Hypothesis testing for moderating effect
The results of the hypothesis testing for the moderating effect are presented in Table 4 and Fig. 2. The results in Fig. 2 show that the moderating link is significant (β= −0.117, P=0.009) but negative, with projects with low adoption of GCPs having the stronger effect relative to projects with high adoption of GCPs (see Fig. 3). This does not support the third hypothesis (H3). The effect size (f 2 ) is interpreted differently as suggested by Kenny (2016). In the analysis of a moderating effect, 0.005, 0.001, and 0.0025 should be used as representative values for small, medium, and large effect sizes (Kenny 2016). Hence, the f 2 for the GCPs*HS➔CS relationship is deemed to be large (> 0.025).

Discussion of results
This study examined the effect of adopting GCPs on CS, the mediating effect of HSP on the relationship between GCPs and CS, and the moderating effect of GCPs on the relationship between HSP and CS. The findings of this study show that GCPs have a significant positive effect on CS which does not support the first hypothesis (H1), HSP partially mediated the relationship between GCPs and CS lending support for the second hypothesis (H2), while it was found that GCPs significantly moderated the relationship between HSP and CS with the stronger effect coming from projects with low GCP adoption which does not support the third hypothesis (H3).
The significant positive effect of GCPs on CS observed in this study can be attributed to the monetary and non-monetary benefits associated with the adoption of GCPs such as better project quality and greater productivity of construction workers (Karakhan and Gambatese 2017). It has been reported that the non-adoption of GCPs could result in fines being imposed on the client and could eventually lead to sealing off of the project site (Onubi et al. 2019), thus leading to extra cost and schedule overruns. Therefore, the positive effect of GCPs on CS could be due to greater economic performance and schedule performance associated with its adoption. Additionally, the positive significant effect of GCPs on CS could be as a result of a situation whereby the perceived benefits associated with GCPs outweigh the cost of investment as opined by Hänninen and Karjaluoto (2017). The compliance with government regulations associated with the adoption of GCPs may also be the reason for the significant positive effect of GCPs on CS obtained in this study. The study's findings differ from those of Venkataraman and Cheng (2018) who reported a general lack of interest from clients on the adoption  Li et al. (2018) attributed this to inefficiencies in the adoption of GCPs, and non-participation of clients in the green construction process. Also, Jiang and Wong (2016) reported a negative significant effect of GCPs on CS which contradicts the findings of this present study. They opined that satisfactory management system in the green construction process, which entails revealing green product information to clients, and communicating effectively and consistently with clients during the construction process, is lacking in green construction projects. This has proven to be responsible for the negative relationship between GCPs and CS (Jiang and Wong 2016). The finding of this study is also in contrast with those of Assaker (2020) and Lee et al. (2018) who reported a no effect and insignificant effect respectively between GCPs and CS. The result of the mediation analysis indicates that HSP partially mediates the relationship between GCPs and CS. This implies that clients can be satisfied with the adoption of GCPs irrespective of the HSP of the projects. In other words, there is a 50% chance of CS with the adoption of GCPs if such practices result in HSP. This finding could be due to other benefits associated with the HSP of projects that adopted GCPs. These benefits include cost savings resulting from fewer accidents, absence of legal disputes, and increased project quality (Howarth and Watson 2010), which could partly lead to CS.
Also, the positive significant mediating effect observed in this study could be due to the improvement in HSP in projects adopting GCPs as reported by Khan et al. (2021). This is similar to the findings of Yacob et al. (2019) and Pham and Kim (2019) who both opined that GCPs lead to CS through HSP. Furthermore, there may have been a change in the way GCPs have been adopted as the adoption of the "prevention through design concept" has become increasingly popular (Zhou et al. 2015). This may have resulted in the safe adoption of GCPs and CS, hence the results obtained in this regard. Additionally, the experience level of the construction workers in GCPs may have increased over the years with its continuous adoption and more training which translates into better HSP to the satisfaction of the clients (Cianciarullo 2019). This study's finding is also similar to the findings of Staniškienė and Stankevičiūtė (2018) and Man et al. (2019) who attributed CS resulting from HSP in projects that adopt GCPs to the importance and value clients attach to their image within the society. The fewer the health and safety issues recorded on a client's site, the greater will be their reputation in the society and their level of satisfaction will be higher. However, the findings of this study are in contrasts with that of Hwang et al. (2018) who opined that the adoption of GCPs is usually characterized by greater safety risks due to the use of unfamiliar construction methods, new technologies,  GCPs green construction practices, CS client satisfaction, HSP health and safety performance hazardous materials, and complexity of the entire construction process. This results in greater construction costs and a lack of CS (Karakhan and Gambatese 2017).
Regarding the moderating effect of GCPs on the relationship between HSP and CS (H3), the result shows that there exists a negative significant moderating effect with the stronger effect emanating from projects with a low GCPs adoption. The discussion and interpretation of the nature of the moderating effect will be done with the aid of Fig 3. Figure 3 shows that the line of the graph for both high GCP adoption and that for low GCP adoption slope upward from left to right. This means that CS increases with an increase in HSP across both sets of projects. Nevertheless, the line of the graph for low GCPs adoption is steeper than that with high GCPs indicating a stronger effect from projects with low GCPs.
The finding regarding the stronger effect of HSP on CS for projects with low GCP adoption conflicts with the findings of Ochieng et al. (2014) who stated that clients take GCPs seriously as a measure of their satisfaction to the extent that they demand it from contractors. This suggests that the level of CS will be higher with a high level of GCPs adoption. Similarly, the results of this study differ from the findings of Shurrab et al. (2019). Shurrab et al. (2019) opined that clients are aware of the advantages of adopting GCPs and are ready to pay for it. By implication, though HSP would lead to CS, CS will be stronger when the level of adoption of GCPs is high relative to when it is low since the clients do not mind the perceived high cost of adoption. Ochieng et al. (2014) also views GCPs as a viable means of improving the project delivery process. All these findings by previous studies are contrary to the findings of this present study. The likely reason for the stronger effect of low GCP adoption on the relationship between HSP and CS obtained in this study could be due to the value clients in the study area attach to GCPs. This could be a problem of value proposition whereby contractors offer to adopt GCPs and the clients may not value it (Lindič and Da Silva 2011). Therefore, even with low adoption of GCPs, the CS would be high since the value they attach to high GCPs is low. Additionally, this finding is similar to that obtained by Issock et al. (2019) who attributed this (stronger effect of HSP on CS for projects with low GCP adoption) to the lack of functional value associated with GCPs which is an important precursor to CS. This study's finding is however similar to that obtained by Lee et al. (2018) in the tourism sector where a negative moderating effect of green practices on CS was reported with low adoption of green practices having a stronger impact on CS than high adoption.

Conclusion
Determining the effect of GCPs on CS and investigating the likely role of HSP in achieving CS on projects that adopt GCPs are of utmost importance in providing the muchneeded guidance to contractors on how to significantly satisfy their clients through deliberate efforts towards HSP when adopting GCPs. As GCPs continue to evolve, there is a growing need to prevent safety issues on construction sites to ensure CS. This study analysed the effect of GCPs on CS, the mediating effect of HSP on the relationship between GCPs and CS, and the moderating effect of GCPs on the relationship between HSP and CS. The results showed that one of the three Fig. 3 Moderating effect of GCPs on the HSP ➔ CS relationship hypotheses (H2) was supported, while H1 and H3 were significant but not supported. The practical contributions and theoretical contributions of this study are presented next.

Practical contributions
The findings of this study make contributions that would be of great benefit to clients, contractors, and other construction industry stakeholders. This study's findings shows that GCPs have a positive effect on CS which is contrary to the results of most previous related studies. Understanding the clients' satisfaction resulting from GCPs may help contractors develop effective strategies such as balancing the adoption of GCPs with adherence to safety regulations, particularly for projects whose clients perceive themselves to be environmentally conscious. The client's assessment of the value they derive from GCPs forms a part of their green construction experience, which consequently influences their satisfaction and continued patronage of the contractors' services. A satisfied client is essential for repeat business and referrals to other prospective clients. In this context, contractors need to reevaluate the role of GCPs and HSP as an integral part of their corporate strategy because of their positive impact on client satisfaction, since CS is a guarantee for their re-engagement in future projects. For instance, contractors could implement GCPs such as energy management practices, waste management practices, and stormwater management practices while adhering to health and safety standards to enhance CS, thus increasing the likelihood of future patronage. The obstacle to this is the inadequate understanding of clients' needs and values by the contractors. Therefore, clients should be carried along throughout the entire construction process for them to fully appreciate the merits of adopting GCPs which would increase their satisfaction. Also, working hand in hand with the clients will enable the contractors to adequately understand client needs and values and allay their fears about any uncertainties surrounding GCPs.
The partial mediating effect of HSP on the relationship between GCPs and CS highlights the all-important role HSP plays in ensuring CS in construction projects. It is recommended that contractors accord great attention to HSP by implementing hazard prevention measures since it serves as a link towards ensuring CS. The finding of this study regarding the mediating role of HSP shows that aside from the indirect role it plays in achieving CS, it could also bring about other deliverables such as lower construction cost and greater quality in project delivery which will ultimately lead to CS. Additionally, the findings of this study would provide an intangible gain to the client's image in the society, which will be aligned with policies like environmental, social, and governance (ESG) criteria and the European Union 2050 climate neutrality guidelines.
Lastly, the negative moderating results obtained imply that the relationship between HSP and CS is stronger when the adoption of GCPs is low relative to projects where the level of GCP adoption is high. This provides crucial guidance to firms with low GCPs adoption that improving HSP through regular training of construction workers on the rudiments of GCPs and provision of adequate personal protective equipment can lead to greater CS. The study's findings also make practical contributions to the body of knowledge by highlighting the importance of contractors to constantly engage with the clients through regular client briefings on work progress and getting the clients acquainted with the properties of green materials and green construction processes while adopting GCPs. This will assist in enlightening the clients on the intricacies involved in green construction, thereby increasing their value for its outcome not minding the cost of adoption. This will go a long way in increasing the CS.

Theoretical contributions
This study makes several theoretical contributions. Although studies such as Karakhan and Gambatese (2017), Zhou et al. (2015), and Dewlaney and Hallowell (2012) examined how HSP can be achieved in projects that adopt GCPs at the design stage, this current study goes further by providing additional insight on the impact of HSP on CS. This study also builds on the study of Hinze et al. (2013) by going beyond how construction HSP can be integrated into green construction projects, to determining how HSP could lead to CS. Additionally, this study contributes to green construction literature by focusing on the satisfaction of the client who is a key stakeholder in the construction industry, thus contributing theoretically to the study of Zhao et al. (2020) who studied the job satisfaction of project managers in green construction projects.

Limitations and future study directions
Though this study has many strengths, there are some limitations that can be addressed in future studies. First, the study was conducted in Nigeria, a developing country in Africa. Due to the restricted geographical nature of the data collected, generalizing the research findings beyond the Nigerian context should be done with caution. However, this study's findings could be relevant to other countries with similar characteristics like Nigeria. It is recommended that similar studies be replicated in other countries and the results compared with those gotten in this study. Secondly, this study adopted the quantitative method using a questionnaire for data collection. As such, the respondents were limited to the options provided in the questionnaire. Future studies can adopt a qualitative research method to obtain more in-depth findings on how to improve HSP on projects that adopt GCPs. Additionally, future studies could consider the importance of early directions during conceptual design to mitigate costs and develop better target alternatives according to the project progress. Only class A contractors were considered in this study; therefore this study's findings should be interpreted within this context. It is recommended that firm size should be considered in future studies to ascertain if the hypothesized relationships could differ depending on firm size or class of contractor.
Lastly, HSP was used as a mediator in this study, and it was found to partially mediate the GCPs-CS relationship. Nevertheless, other factors such as economic performance and schedule performance could be considered as mediators in future studies in addition to HSP to obtain more insights.
Author contribution All authors contributed to the study conception and design. Data collection and analysis were performed by HOO, NY, and ASH. The first draft of the manuscript was written by HOO, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations
Ethics approval and consent to participate This study involved human participants. Ethics approval was granted by the Ethics Committee for Research Involving Human Subjects, Universiti Sains Malaysia (JEPeMUSM), with study protocol code USM/JEPeM/18090438 on 9 May 2019.

Consent for publication Not applicable
Competing interests The authors declare no competing interests.