The Influence of Employee Engagement, Work Environment and Job Satisfaction on Organizational Commitment and Performance of Employees: A Sampling Weights in PLS path Modelling

Design/MethodologyData were obtained from seven hundred and twenty (720) employees from selected financial banks in the Greater Accra Region of Ghana through simple random probability sampling. However, seven hundred (700) responses were deemed accurate and therefore used in the analysis. In the analytical process, Weighted Partial Least Squares (WPLS) and Partial Least Squares (PLS) based on Structural Equation Modeling (SEM) were employed.


Introduction
In light of challenging economic circumstances, any establishment's sustainable growth relies on workers' increased performance. Employee performance (EP) is the successful and reliable execution of the duty in a demanding environment using the stipulated period's tools. EP is linked to the activities carried out to achieve organizations' goals and objectives (Motowidlo, Lievens, & Ghosh, 2018). The organization's creation is primarily focused on the EP as it affects efficiency (Chikazhe, Makanyeza, & Kakava, 2020), and lack of required productivity affects the organization's sustainability (Anyakoha, 2019). EP is critical because it generates optimal organizational efficiency (Abdirahman, 2018). Increased level of dedication serves as the driving force behind EP. When evaluating the impact of organizational commitment (OC), several studies reinforce the belief that OC inspires workers to perform to the highest efficiency standards (Berberoglu, 2015).
The importance of the results of the study cannot be devalued. First, this study's findings would allow companies to build practical human capital management approaches to improve their corporations' overall value. Next, the results of this study will encourage decision-makers to create tailored strategies and initiatives that will positively inspire the development and survival of organizations around the world. Third, the research's results could be used by controlling bodies such as the Bank of Ghana and other financial bodies to advance their regulatory structure further. Finally, the research provides more modernized scientific data to current human resource management literature in Ghana concerning OC and EP. This is of immense value to the academic field, as it serves as reference material for students and scholars who may wish to do further exploration on the current topic.
Several antecedents of employee performance have been examined and investigated by prior studies. For example, leadership practices (Le & Tran, 2020); motivation (Ackah, 2014); training and development (Ampomah, 2016;Boadu, Dwomo-Fokuo, Boakye, & Kwaning, 2014); leaders' behavior (Obuobisa-Darko, 2019) and communication (Otoo, 2016). However, very little has been done on variables such as employee engagement (EE), work environment (WE), and OC in the Ghanaian context. This research was conducted to aid fill this gap. Studies on EP, OC, job satisfaction (JS), EE, and WE are abundant. The conclusions are, however, inconsistent. For, e.g., (Cesário & Chambel, 2017) researched EP, OC, and EE in Portugal. As a result, there was a positive correlation between EP and OC and between EE and EP. Eliyana and Ma'arif (2019) have discussed the relationships between EP, JS, and OC in Indonesia. It was reported that there was a strong association between JS and EP. There was, however, a negative correlation between OC and EP. Abdirahman (2018) examined the influence of JS and OC on EP in Malaysia. The findings indicated a positive relationship between JS and EP and amid OC and EP. In the public sector of Ghana, Amoako-Asiedu and Obuobisa-Darko (2017) examined the interrelationship between EE and EP. A significant linkage between EE and EP was obtained from the findings. Ahakwa, Yang, Agba Tackie, Afotey Odai, and Dartey (2021) delved into the relationship between WE and OC among MMDAs employees in Ghana. From the finding, there was a significant positive connection between the two variables. The studies as mentioned above, among others, are deficient in scope as most human resource management studies related to OC and EP in Ghana are more associated with organizations operating in the primary and industrial sectors, to the detriment of those operating in the service field, to the best of our knowledge. By focusing on only accredited banking organizations in Ghana, this study contributively fills that gap. The research adds in the following ways to the current body of literature: First, most previous studies studied the relations between OC, EE, WE, JS, and EP using partial least squares focused on structural equation modeling (SEM). Most studies, however, neglect the Weighted PLS-SEM (WPLS-SEM). WPLS-SEM uses weighed correlations and weighted regression results to estimate the PLS path model (Becker & Ismail, 2016). The WPLS-SEM permits investigators to specify a weighting vector that determines the importance of each observation's results. This study employed WPLS-SEM and compared the result with the original PLS-SEM. This study also used the current method of assessing the model's predictive relevance, "the PLSpredict" proposed by (Hair et al., 2020;Shmueli et al., 2019), which is not common in existing studies.
Second, numerous studies of OC, EE, WE, JS, and EP have been done. However, only a small number of those studies considered the issue of common method bias (CMB). According to Kock (2015), a probable cause of CMB is the implicit social desirability allied with answering questions in a questionnaire in a particular way, again causing the indicators to share a certain amount of common variation. To help address CMB's issue, the researchers provided appropriate reliability evidence, factor structures, and convergent and discriminant validity as suggested by (Conway & Lance, 2010). A full collinearity assessment approach proposed by Kock (2015) was also employed to deal with CMB's issue.
Third, numerous studies of OC, EE, WE, JS, and EP have been done. However, only a small number of those studies considered either mediation or moderation in their analysis model. To the best of our knowledge, this is the first study to report on this study's variables in moderated-mediated interaction in the Ghanaian context upon an extant review of the literature. Therefore, this research aims to fill the gap by creating an integrative model that considers many variables and mechanisms relevant to EP, a moderated-mediated model.
Last, most existing studies end up using a small sample size for their analysis when using PLS-SEM. According to Fornell and Bookstein (1982), when models contain several constructions and a wide number of objects, PLS-SEM proposes better solutions with small sample sizes. Hair Jr, Hult, Ringle, and Sarstedt (2016) suggest that "some researchers have wrongly and misleadingly used these features to produce studies with exceedingly small sample sizes, even when the population is large and accessible without much effort." Unlike other studies, this research used a larger sample size of seven hundred (700) for the study's analysis. Such a larger sample size provides more accurate mean values and gives a smaller margin of error. Wamba et al. (2017) indicated that "a greater number of prior studies on sample size requirements in PLS-SEM unnoticed the fact that the process also proves valuable for evaluating large data quantities." The contributions mentioned above are novel since they are deficient in OC and EP studies undertaken in the Ghanaian context. The study is eventually unique since the investigators themselves carried it out; the study's hypothesis and intent are clearly defined; the techniques used are fully detailed; the findings are properly represented, and the policy implications are properly explained. The rest of the report is organized as follows: the "Review of Related Literature" section presents the literature promoting the subject under study. At the same time, the "Method" part reflects the study technique. Empirical findings of the analysis are summarized in the "Empirical Results" section, while discussions, practical implications, and conclusions are the final section of the research.

Review of Related Literature Employee Engagement
EE is an important term in an organization, which has gained substantial interest in scholarly study. Saks (2019) referred to the EE as the degree that one is conscientious and involved in his/her work roles. Also, "EE is perceived as a good and satisfactory behavior associated with work that is marked by three components: vigor, interest and devotion" (Rothmann, 2017). "An engaged employee is projected to experience traits such as socially, psychologically and cognitively" (W. Kim, Khan, Wood, & Mahmood, 2016). Kang and Sung (2017) described EE as "the degree of participation, communication, intimacy and impact of an employee with a specific brand, the involvement of an employee with a brand, irrespective of the medium where they make the decisions." Furthermore, (Hanaysha, 2016a) conceived of EE as the productive, interpersonal, emotional conduct of work, motivating workers to mentally, cognitively, and physically express and plan for their duties. To obtain valuable market success results for various organizations, the EE, according to Rothmann (2017), is extremely significant. The authors have shown that it is important for companies to involve their workers, as EE establishments have a greater degree of customer satisfaction and commitment, extra efficient, and profitable than those with less EE (Rothmann, 2017). Hanaysha (2016a) has recommended cultivating EE's idea as retrenchment reduces workers' motivation and devotion to their establishments. Therefore, low amounts of dedication have a detrimental impact on the OC and retaining of workers. Jiony, Tanakinjal, Gom, and Siganul (2015) suggested that a well-performing company is dependent on its ability to maintain a safe, engaged and dedicated workforce through interaction.

Organizational Commitment
It is undoubtedly very required for all establishments to cultivate OC since personnel is the key foundation of sustained achievement and efficiency. "OC is described as an emotional attitude that binds personnel to an establishment in a way that decreases turnover intention" (Ahakwa, Yang, Agba Tackie, et al., 2021). Lee, Ashford, Walsh, and Mowday (1992) proposed the most generally accepted definition for OC as "the level to which a person's participation in his organization." The authors added that loyalty is demonstrated by an employee's ability to labor successfully in an establishment and the desire to sustain the relationship devoid of attempting to turn to another (Lee et al., 1992). Organizations with greatly committed personnel, since it's widely agreed that OC might lead to countless organizational results; reduced turnover, greater motivation level, enriched citizenship conduct, and continuous organizational support (Ahakwa, Yang, Agba Tackie, et al., 2021). "The commitment of workers is a sign of greater devotion and improved efficiency" (Porter, Steers, Mowday, & Boulian, 1974). Committed personnel often work assiduously to fulfill objectives of establishment and appear to positively consent to their values (C. S. Cheah, Chong, Yeo, & Pee, 2016). Many positive behavioral outcomes can be correlated with OC of employees, such as greater retention of workers, motivation, efficiency, quality of work, and willingness to make sacrifices to enhance the reputation and performance of organizations (Somers, 1995). OC is a crucial element in assessing organizations' effectiveness, which increases employee morale and EE (Hayat, Azeem, Nawaz, Humayon, & Ahmed, 2019;Hendri, 2019;Yousef, 2017). OC often has a clear correlation with the actions and performance of employees. If an employee has an OC, there will be fewer chances for absence and turnover (R. Ahmad, Islam, & Saleem, 2019;Igbaria & Greenhaus, 1992;Joe-Akunne & Ezeh, 2019;Karunarathne & Wickramasekara, 2020). Therefore, it is crucial to regularly review employees' commitment to resolving any problems that may occur and ensure that workers maintain a good attitude to work, which is indispensable to overall organizational success.

Work Environment
The WE is a significant element that influences the JS and OC of employees to the organization. The WE refers to the surrounding of an establishment where employees do their work. "The WE is linked to the atmosphere of a specific company in which its workers conduct their duties" (Danish, Ramzan, & Ahmad, 2013). Undoubtedly, since their needs are likely to be fulfilled, a facilitative and healthy work atmosphere will attract employees. To succeed, companies should design their WE to enhance employees' dedication and motivation that eventually contribute to promising results. A good WE include all the essentials of a job, such as the amenities to perform responsibilities, a relaxed workspace, protection, and no noise. Hanaysha (2016a) found that, relative to those who feel insecure, workers who feel relaxed with their WE are likely to work more efficiently and enjoy the working process. Managers should also strengthen the elements of the WE to make certain of the well-being of their workforces. Past research demonstrates that WE can be measured in many respects. Ahakwa, Yang, Agba Tackie, et al. (2021) indicated that "the WE involves elements such as participation; group cohesion; provision for supervisors; role direction; self-sufficiency; clarity; creativity, physical well-being, and management power." Also, Ahakwa, Yang, Agba Tackie, et al. (2021) defined a range of dimensions for assessing the WE, including: "job challenge, job autonomy, leader concern, and support, leader work facilitation, working group cooperation, workgroup spirit, position uncertainty, fairness and reward system equity. Therefore, the WE may be evaluated in terms of any factor that influences the actions of an employee in his or her organization.

Employee Job Satisfaction
JS is the wonderful psychological state arising from the enjoyment of a person's own work experience (Liu, Aungsuroch, & Yunibhand, 2016). It could also be perceived as workers' mindset towards their employers, the environment of corporate, social, and physical work, and the benefits received (Yousef, 2017). JS implies how a member of the organization feels about work (Qureshi & Hamid, 2017). Such emotions may be positive or negative; more positive feelings indicate JS's degree is high. JS also defines a worker's optimistic feelings about the workplace. Judge and Locke (1993) showed a strong correlation between the features of the work and people's desires. There is also a consensus among scholars that Maslow's theory of needs clarifies this association between work characteristics and human needs. Stajkovic and Luthans (1998) suggested that JS has three components; first, JS refers to an employee's emotional reaction to WE. Second, JS may be calculated by estimating how well results fulfill requirements. Last, JS can be assessed by many behaviors relevant to work. The success management framework often stresses JS employees (N. Ahmad, Iqbal, Javed, & Hamad, 2014). JS is to build optimistic feelings among employees regarding jobs Robbin and Judge's (2008). Greater JS produces more optimistic thoughts about their jobs in the minds of workers. Badran and Youssef-Morgan (2015) found out that JS induces optimistic emotional feelings arising from job appraisal.

Employee Performance
Performance is defined as the product of trained employees in some particular circumstances (Obicci, 2015). EP is the product or degree of an employee as a whole's progress in executing the task over a given amount of time relative to other things, such as the quality of work, objective, or standards that have been previously defined and collectively decided upon (Obicci, 2015). Vrinda and Jacob (2015) observed a dispute between the personal life and performance of workers. Dahie, Takow, Nur, and Osman (2016) examined that performance effectively is the product of work with a fair corporate responsibility without interrupting any regulations and organizational objectives. Darma and Supriyanto (2017) say that EP results from perfection and everyone's quantity in directing his/her job obligations.

Hypotheses Development and Conceptual Framework Employee Engagement and Organizational Commitment
Previous findings have uncovered that EE has a significant positive influence on OC (Hanaysha, 2016a;Imam & Shafique, 2014;Nazir & Islam, 2017). Engaged workers make greater attempts to work assiduously, are highly likely to drive further than their necessary and anticipated number of work assignments (Lockwood, 2007). Also, engaged workers tend to find their working conditions and workplace principles positively impact their physical and psychological security at work (Agyemang & Ofei, 2013). Schaufeli (2013) previous research has indicated that EE influences the degree of OC. H1: Employee engagement positively influences organizational commitment.

Employee Engagement and Employee performance
Past studies have made known that EE has a significant positive influence on EP (Anitha, 2014;Ayub & Islam, 2018;Sendawula, Kimuli, Bananuka, & Muganga, 2018). However, Kuruppuge and Gregar (2017) identified a negative relationship between EE and EP. The level of loyalty strongly influences the degree of EP that an individual has to his company and its beliefs (Sendawula et al., 2018). A dedicated worker is mindful of the corporate setting and partners with peers to enhance job efficiency for the organization's benefit (Anindita & Seda, 2018). EE is regarded as capacity, interest, participation, effectiveness, vigor, motivation, excitement, and a positive state (Men, O'Neil, & Ewing, 2020). Truss, Shantz, Soane, Alfes, and Delbridge (2013), revealed that "committed workers have a healthy mindset and a work-based state of mind marked by vigor, commitment, and interest, making staff mentally present at work, reducing their tendency to make mistakes and errors related to work." Employees who are engaged are more prepared to know new things, according to (Sugianingrat et al., 2019). This demonstrates that engaged employees are able to put their ideas into motion and, as a result, reach high levels of success at work (Sugianingrat et al., 2019). Kruse (2012) examined 28 research studies by various scholars and discovered a connection between EE and operation, revenue, pricing, safety, productivity, earnings, and overall profitability. Monica and Krishnaveni (2018), who expressed that engaged workers frequently portray a strong positive emotional bond with their job and are frequently more active, efficient, stronger, happier, and less likely to abandon their employer, all support the correlation between EE and EP. This result is also in line with Dhir and Shukla (2018), who performed a meta-analysis and discovered that EE is linked to higher performance, consumer retention, and therefore customer loyalty. Employees who are engaged outperform their discontented colleagues (Shuck, Reio Jr, & Rocco, 2011).
H2: Employee engagement positively influences employee performance.

Work Environment and Organizational Commitment
A variety of studies have found that the WE has had a major positive impact on OC Hanaysha, 2016a;Khuong & Le Vu, 2014). Pitaloka and Sofia (2014) establish a positive impact on JS and OC in a conducive WE. Haggins (2011) established that WE played an important role in persuading OC. By Giffords (2009), one of the key contributors to OC is the WE. According to Rayton (2006), a safe working atmosphere, clear connectivity, and an adequate workload are deciding factors in an employee's OC. If this aspect is lacking, people no longer feel at ease; they only come to work and work when their thoughts are elsewhere; they have no compelling excuse to remain and survive in the organization. Employees who work in a pleasant environment feel more committed, according to (Zainudin, Rashid, Murugeesan, Che Zainal, & Malek, 2019). Also, a comfortable WE will lead to increased employee commitment (Nwachukwu, Ezeh, Ogochukwu, Nkechinyere, & Dumle, 2019). Among bankers, Karacsony (2019) discovered a negative connection between employee environment stressors and organizational engagement. According to Hanaysha (2016a), the WE has a significant impact on OC. According to Hanaysha (2016a), the workplace atmosphere tailored to the workforce demonstrates commitment on the part of the company.
H3: Work environment positively influences organizational commitment.

Work Environment and employee performance
Past studies have shown that WE have a significant positive effect on EP (Badrianto & Ekhsan, 2020;Imran, Fatima, Zaheer, Yousaf, & Batool, 2012;Nguyen, Dang, & Nguyen, 2015;Rorong, 2016). According to E. M. Putri, Ekowati, Supriyanto, and Mukaffi (2019), WE are among the variables that impact EP. The study conducted by E. M. Putri et al. (2019) indicated that the agency's WE could positively and significantly affect EP. According to Rorong (2016), the office atmosphere's practical décor and design eventually helped enhance workers' experiences and necessitate improved efficiency. Previous researchers stated that the physical working environment helped to deter employees' ability to connect with their work roles and influence their behavior (Al-Omari & Okasheh, 2017). A study had shown that a conducive physical working environment could reduce absenteeism and enhance employee performance (Chandrasekar, 2011). Thus, to retain employee performance, the organization had to improve the physical working environment. Nematchoua, Ricciardi, Orosa, Asadi, and Choudhary (2019), revealed that a suitable workplace temperature energizes an office occupier to work at the employee's best.
H4: Work environment positively influences employee performance.

Organizational commitment and employee performance
The OC and EP relationship is confirmed by historical research. Hidayah and Tobing (2018) clarify that OC affects EP. The findings suggest that OC types, such as continuous, normative, and affective, are related to employee job efficiency. Susanty and Miradipta (2013) reveal the effect of OC on EP. The results showed that OC forecasts EP separately and jointly. The studies carried out on university teachers have concluded that OC has a favorable association with EP. The OC and EP are being studied by Suliman and Iles (2000) in three industrial units. This research has demonstrated that organizational participation is a three-dimensional phenomenon and has a beneficial association with job efficiency.
Another research indicates that OC has a favorable association with work efficiency. Dixit and Bhati (2012) observed that employee motivation is an essential concern because it can forecast EP, absenteeism, and other activities. Employees with a high level of organizational responsibility are more efficient and profitable, which benefits the organization as a whole (Paramita, Lumbanraja, & Absah, 2020). OC has a significant and constructive association with performance. According to Ramli (2019) organizational commitment is an antecedent that can function to determine job performance. According to (Paramita et al., 2020), OC is linked to EP. According to Singh (2019), OC has a favorable relationship with overall job results. Barrick, Stewart, and Piotrowski (2002) discussed and identified a constructive association and a precise effect between OC and EP. Fu and Deshpande (2014) discovered that OC has an important and beneficial impact on EP. Similarly, Jamal (2011) demonstrates that OC has a direct impact on employee success. H5: Organizational commitment positively influence employee performance. Chaudhry, Jariko, Mushtaque, Mahesar, and Ghani (2017) suggested that satisfaction and EP are correlated to one another, and the outcome of work efficiency is satisfaction. Platis, Reklitis, and Zimeras (2015) analyzed employee satisfaction and efficiency and identified that work satisfaction provides workers with feedback for improved performance. Better worker efficiency is the degree of employee satisfaction (Shah & Jumani, 2015). The significant EP metrics at the recruiting level were studied by (Muntazeri & Indrayanto, 2018). They concluded the degree of JS and motivation influences the employee's productivity. The low level of JS negatively impacts employee motivation and sequentially affects the accomplishment of corporate goals and results (NATH & Agrawal, 2015) H6: Job satisfaction positively impacts employee performance.

The mediation role of organizational commitment
Previous studies have linked EE and OC (Anindita & Seda, 2018;Hanaysha, 2016a;Khalid & Khalid, 2015;W. H. Putri & Setianan, 2019) and WE to OC (Badrianto & Ekhsan, 2020;Imran et al., 2012;Nguyen et al., 2015;Rorong, 2016). Also, OC is a critical factor in influencing EP, resulting in high organizational performance. Cesário and Chambel (2017) have linked OC and EP and found a significant link between the two variables. Yuniarti and Prasetyaningtyas (2020), through OC, there is a positive connection between EE and EP. This is because dedicated workers feel positive feelings that extend their thought, allowing them to become more attentive and immersed in their job. Building on these, the present study theorizes that EE and WE can be associated with higher OC levels, resulting in a higher level of EP.
H7: Organizational commitment mediates the connection between employee engagement and employee performance.
H8: Organizational commitment mediates the link between work environment and employee performance.
H9: Job satisfaction moderates the association amid employee engagement and organizational commitment.
H10: Job satisfaction moderates the nexus amid work environment and organizational commitment. H11: Organizational commitment will mediate the relationship between employee engagement and employee performance, and the association will be stronger when job satisfaction is high. H12: Organizational commitment will mediate the link between work environment and employee performance, and the association will be stronger when job satisfaction is high.

The Conceptual Framework
Based on the literature reviewed, the conceptual framework for this study is presented below.

Research Design
A research design is a collection of procedures and techniques used to capture and analyze the measurements of factors recognized in a research problem (Ahakwa, Yang, Tackie, & Bankole, 2021). To evaluate classification features, quantify numbers, and construct a predictive pattern to test hypotheses and explain results, the research used quantitative approaches.

Target Population
The population is defined as a category of the research object, has one or more common characteristics, and is the subject of interest to the investigator. Employees and supervisors from Ghana Commercial Banks (GCB), Ecobank Ghana (EBG), Zenith Bank of Ghana, and Agricultural Development Bank (ADB) in the Greater Accra Metropolis were used as the study's population.

Research sampling and sample size.
The sampling applies to an approach to choosing a part of the sample population that will represent the whole study population. This research followed a simple random sampling technique where every employee was given an equal opportunity to answer the questions asked. The sample size used for the analysis was seven hundred (700), all of whom were employees from the Greater Accra Metropolis banking sector.

Data instrument and collection
To achieve the objectives of the study, the researcher gathered information from the study population. The research used the questionnaire as a medium to request information from the population. The questionnaire included questions related to the participants' demographic characteristics. There were questions in the second part of the questionnaire that helped examine the variables to be evaluated. Participants were called upon to rate the items based on the 5-point Likert with the scaling pole ranging from strongly disagree (1) to strongly agree (5). Seven hundred and twenty (720) online questionnaires were sent to respondents through various digital platforms. Seven hundred (700) were deemed fit and accurate, then used for discussion.

Measurement of Variables
OC was measured with eight items adapted from (Allen & Meyer, 1996), and EP was measured with six items taken from (Cropanzano, Rupp, & Byrne, 2003;Lutwama, 2011). Also, EE and WE were measured with four items, each adapted from (Hanaysha, 2016). Last, JS was also measured with four items adapted from (Clack 2020). All the items were measured on a scale of 1 to 5. Table 1 reveals the items used for each variable. I focused on the positive aspects of a work situation instead of on the negative aspects.

OC1
I would be very happy to spend the rest of my career in this organization. OC2 I do not feel a strong sense of belonging to this organization. OC3 It would be very hard for me to leave my job at this organization. OC4 I am proud to tell others that I am part of this institution OC5 I am willing to put in high effort to help this institution be successful. OC6 This organization deserves my loyalty.

OC7
For me, this is one of the best institutions for which to work. OC8 I would feel guilty if I left this organization now

Data Analysis
The research involved exploratory and confirmatory studies in confirming the validity of the model. SPSS version 26.0 was used to process descriptive statistics to measure the demographic profile of the samples. Partial Least Squares (PLS) analysis using SmartPLS 3.0 software was used to evaluate the research model. The measurement model was tested for validity and reliability of the constructs using both the PLS-SEM algorithm and WPLS-SEM algorithm. The structural model was then analyzed in conjunction with the two-stage analytical procedures suggested for SEM (Hair Jr et al., 2016). We used both PLS-SEM and WPLS-SEM to looked at the R², path coefficient (β), and corresponding t-statistics through the 5000 resample bootstrapping process suggested by Hair Jr et al. (2016). They also recommended that researchers report on predictive significance (Q²) and effect sizes (f²) together with the basic measures. We also looked at assessing the prediction error degree using PLSpredict as recommended by (Shmueli et al., 2019).   (59). This indicates that more females responded to the items than males. The finding, therefore, reveals that more females than males dominate the banking sector of Ghana. Hence researchers' decided to apply weight to each observation to have an accurate representative sample.  Age and gender distribution of the population of IndexMundi on the Ghana Demographics Profile (www.indexmundi.com) was collected to compute each observation's sampling weight. We first separately used the populations of the age group and gender parted by the population to measure the proportion of the population. Next, we used each group sample separated by the sample's sum to produce 'Sample Proportion.' Finally, we obtained the 'Weight' group findings on the 'Population Proportion' being split by the 'Sample Proportion.' Table 3 compares the age and gender survey and demographic distributions. Next, using the values described in Table3, we produced a weighting variable. For example, Table 4 uses the values described in Table Table 5 displays the reflective measurement model for both PLS-SEM and WPLS-SEM for internal consistency reliability, convergent, and discriminant validity. The indicator loadings, average variance extracted (AVE), composite reliability (CR), and Cronbach alpha (CA) values of the latent constructs were extracted after performing the confirmatory factor analysis on all the constructs. In both PLS-SEM and WPLS-SEM models, the indicator loadings were above 0.6 as recommended by (Chin, Peterson, & Brown, 2008). The indicator loadings for both PLS-SEM and WPLS-SEM ranged from 0.616-0.953 and 0.638-0.960, respectively. The Cronbach alpha for WPLS-SEM ranged from 0.773-0.939, and for PLS-SEM ranged from 0.763-0.963; thus, both PLS-SEM and WPLS-SEM met the suggested threshold of above 0.7 (Hair Jr et al., 2016). Moreover, we test for the CR of all the constructs, and the value ranged from 0.796-0.948, exceeding the suggested figure of 0.7 or greater (Hair Jr et al., 2016), with OC bearing the highest value in the WPLS-SEM. In the PLS-SEM model, CR ranged from 0.846-0.969, exceeding the suggested figure of 0.7 or greater (Hair Jr et al., 2016), with OC bearing the highest value. All the constructs in both PLS-SEM and WPLS-SEM met the AVE's minimum required value of 0.5, as suggested by (Hair Jr et al., 2016 (Hair Jr et al., 2016) and (Chin et al., 2008) for the internal consistency and reliability checks, we can conclude that the model is accurate enough for the analysis in both WPLS-SEM and PLS-SEM.  Tables 6 and 7 present the VIFs (outer values and inner values) assessments of the various constructs' collinearity values. To better understand collinearity in the statistical model, Dormann et al. (2013) state that it is used to estimate the linkage amid a dependent variable and a group of independent (predictor) variables. Grewal, Cote, and Baumgartner (2004) argue that the collinearity problem among variables may occur just by chance, especially when the sample size used for the analysis is low. However, perfect collinearity can arise if all the variables used in the study are of the same linear qualities (Dormann et al., 2013). According to J. H. Kim (2019), if the VIF figures are more than 5, there is a collinearity problem in the model. However, if the VIF figures are lower than 5, then the model is free from the problem of collinearity. Our VIF assessment for both PLS-SEM and WPLS-SEM revealed values less than five (5), suggesting no collinearity problems in the model (Tackie et al., 2020).

Empirical Results
Also, the occurrence of a VIF greater than 3.3 is proposed as an indication of pathological collinearity and indicates that the model may be contaminated by common method bias. Therefore, if all VIFs resulting from a full collinearity test are equal to or lower than 3.3, the model can be considered free from common method bias (Kock, 2015). Hence our VIFs (outer and inner values) in both PLS-SEM and WPLS-SEM as presented in Tables 6 and 7 suggest that our model is free from CMB. To weigh the discriminant validity, which represents the degree to which the measures are not replicating some other variables, low correlations between the measure of interest and other constructs' measures are indicated. Table 8 shows that each construct's AVE square root (diagonal values) is greater than its corresponding correlation coefficients, suggesting sufficient discriminant validity (Fornell & Larcker, 1981) in the PLS-SEM and WPLS-SEM. We also used the Heterotrait-Monotrait (HTMT) criterion to measure discriminant validity. Compared with Fornell-Larcker's Criterion, the HTMT gives a more rigorous outcome. (Henseler, Ringle, & Sarstedt, 2015) debunked Fornell-Larcker Criterion because the researchers believed that it is not reliable enough to distinguish lack of discriminant validity in ordinary research. Therefore, a multitrait-multimethod matrix analysis tool; thus, the HTMT ratio of correlations is considered more reliable. Table 9 gives the values for the discriminant validity measured using the Henseler et al. (2015) alternative approach. According to (Kline, 2011) when the HTMT value is larger than the threshold figure of 0.85, there is a problem with discriminant validity. However, as presented in Table 9, the discriminant figures for all the constructs (EE, WE, JS, OC, and EP) were below the HTMT threshold value of 0.85 in the PLS-SEM and WPLS-SEM. Therefore, using the Heterotrait-Monotrait criterion, our analysis is free from discriminant validity problems (Kline, 2011) in PLS-SEM and WPLS-SEM.  We assessed the direct relationships in the study's model. Table 10 reveals, first H1 [EE → OC] in the PLS-SEM and WPLS-SEM; is supported (β=0.148; t-value=4.170, p < 0.01) and (β=0.214; t-value=4.285, p < 0.01) respectively. This finding indicates that EE positively and significantly influence OC. Second, H2 [EE → EP] in the PLS-SEM and WPLS-SEM; is supported (β=0.321; t-value=8.906, p < 0.01) and (β=0.298; t-value=5.406, p < 0.01) respectively. This finding implies that EE positively and significantly influence EP. Third, H3 [WE → OC] in the WPLS-SEM and PLS-SEM; is supported (β=0.243; t-value=3.951, p < 0.01) and (β=0.143; t-value=4.362, p < 0.01) respectively. This finding denotes that WE positively and significantly influence OC. Fourth, H4 [WE → EP] in the WPLS-SEM is unsupported (β=0.131; t-value=1.538, p < 0.05), however, H4 [WE → EP] in the PLS-SEM is supported (β=0.174; t-value=5.387, p < 0.01). This finding suggests that WE positively and significantly influence EP in the PLS-SEM but does not in WPLS-SEM model. Fifth, H5 [OC → EP] in the PLS-SEM and WPLS-SEM; is supported (β=0.211; t-value=4.454, p < 0.01) and (β=0.289; t-value=4.353, p < 0.01) respectively. This finding infers that OC positively and significantly influence EP. Last, H6 [JS → EP] in the PLS-SEM and WPLS-SEM; is supported (β=0.338; t-value=10.134, p < 0.01) and (β=0.316; t-value=4.647, p < 0.01) respectively. This finding reveals that JS positively and significantly influence EP.

Assessment of the Structural Model
Comparing the two models (WPLS-SEM and PLS-SEM), we discovered three path coefficients varying more than 0.05. First, in the WPLS-SEM, the path coefficient for EE → OC is 0.214, and, in the PLS-SEM, the path coefficient is 0.148. The variation is 0.066. Second, the path coefficient for WE → OC varies by 0.100 amid the two models, with the WPLS-SEM path equal to 0.243 and that of PLS-SEM is 0.143. Third, the path coefficient for OC → EP in the WPLS-SEM is 0.289, and that of the PLS-SEM model is 0.211. The difference is 0.078. Finally, these relationships were all significant at 1% level in both the PLS-SEM and WPLS-SEM. The WPLS-SEM model has a larger path coefficient (β) values than the PLS-SEM model, making the WPLS-SEM more preferred. However, all direct relationship in the PLS-SEM model was supported. We assessed the indirect effect; thus, mediation analysis. First, table 11 reveals that H7 [EE → OC → EP] in the PLS-SEM and WPLS-SEM; is supported (β=0.031; t-value=3.219, p < 0.01) and (β=0.062; t-value=2.945, p < 0.01) respectively. This finding indicates that OC positively and significantly mediates the association between EE and EP. Second, H8 [WE → OC → EP] in the PLS-SEM and WPLS-SEM; is supported (β=0.030; t-value=2.815, p < 0.01) and (β=0.070; t-value=2.608, p < 0.01) respectively. This finding implies that OC positively and significantly mediates the link between WE and EP.
Comparing the two models (WPLS-SEM and PLS-SEM). First, in the WPLS-SEM, the path coefficient for EE → OC → EP is 0.062 and, in the PLS-SEM, the path coefficient is 0.031. The variation is 0.031. Second, the path coefficient for WE → OC → EP varies by 0.040 amid the two models, with the WPLS-SEM path equal to 0.070 and that of PLS-SEM is 0.030. Also, OC fully mediates WE → EP in the WPLS-SEM model as compared to PLS-SEM with partial mediation. Finally, these relationships were all significant at 1% level in both the PLS-SEM and WPLS-SEM. The WPLS-SEM model has larger path coefficient (β) values and better mediates relationships than the PLS-SEM model, making the WPLS-SEM more preferred.     Effect size measures the exogenous latent construct on the endogenous construct using the f². Cohen (1988) guidelines were used to measure the effect size, which are 0.02 for small effects, 0.15 for medium effects, and 0.35 for large effects. Table 13 revealed that all relationships had a medium effect in the WPLS-SEM. All relationships had a medium effect in the PLS-SEM except [JS*EE→OC] and [JS*WE→OC] with small effects. The exogenous latent construct in the WPLS-SEM better influences the endogenous construct than that of the PLS-SEM.  Table 14 discloses the R², which is the total amount of explained variance in the endogenous constructs'. With the WPLS-SEM, the R² figure of EP is 0.423, whereas the R² figure of OC is 0.506. In the PLS-SEM, the R² of EP and OC are 0.451 and 0.501, respectively. These outcomes signpost that when relating the two models (PLS-SEM and WPLS-SEM), the R² of both endogenous constructs does not vary much in size. Table 14 reveals the blindfolding procedure, which determines the predictive significance Q² of PLS path models (Hair Jr et al., 2016). We acquired our Q² figure through cross-validated redundancy procedures. A Q² figure bigger than zero means that the model has predictive relevance; however, a Q² figure beneath zero indicates the model's predictive significance lacks. For this assessment criterion, we also discovered a variation of more than 0.05 in the two models. In the PLS-SEM and WPLS-SEM, the Q² figure of EP is 0.261 and 0.193, respectively, leading to a variation of 0.068. The variation amid the two model's Q² figures of OC is 0.073. With the WPLS-SEM, the Q² figure of OC is 0.304, while with PLS-SEM, the Q² figure of OC is 0.377. Both PLS-SEM and WPLS-SEM model has predictive relevance; however, PLS-SEM better predicts the model's relevance. We performed "PLSpredict" on the two models (Shmueli et al., 2019). Table 15 revealed that for both OC and EP constructs, the PLS-SEM study resulted in marginally lower prediction errors for both targets constructs given MAE and RMSE and high Q² predict figure. Generally, our PLSpredict test showed that the model calculated using PLS-SEM matched the data better and realized a higher predictive power.

Discussion
Our study support and advance previous studies wherein EE significantly impact OC (Hanaysha, 2016a;Imam & Shafique, 2014;Nazir & Islam, 2017) and EP (Anitha, 2014;Ayub & Islam, 2018;Sendawula et al., 2018). This outcome proposes that the more workers are engaged in the workplace, their commitment and performance to the establishment will be high. An employee who displays good working behavior by engagement is likely to express greater commitment and success to the company due to the great zeal and bravery for accomplishment, which justifies this study's finding (Schaufeli & Bakker, 2004). The author further indicated that engaged workers appear to encourage meaningful interactions and efficiency in their organizations. Also, WE positively influence OC Hanaysha, 2016a;Khuong & Le Vu, 2014) and EP (Badrianto & Ekhsan, 2020;Imran et al., 2012;Nguyen et al., 2015;Rorong, 2016). This finding discloses that WE are a critical element that can impact employees' commitment and performance in an organization. Therefore, this result's practical implication suggests that authorities in charge of organizations have to be mindful of the value of creating a conducive environment by enhancing OC among their workforces, therefore leading to increased performance. For example, the provision of leisure facilities and preserving a green and sanitary environment can play a central role in coaxing employees' actions. Moreover, the outline of the workplace and organizational philosophy is also fundamental to enhance OC and performance.
Furthermore, JS positively contributes to previous studies by impacting EP (Chaudhry et al., 2017;Platis et al., 2015), and OC also positively influence EP (Hidayah & Tobing, 2018;Susanty & Miradipta, 2013). This result means that the presence of JS and OC in an establishment leads to improved EP. Individuals with a high degree of company loyalty and happiness can reflect the organization's good behavior, offer the best of their abilities, sacrifice, be faithful to the organization and still have a desire to stick with it. This implies that an individual with a high degree of OC and satisfaction aims to demonstrate high level of work performance. On the other hand, an individual with low level of organizational commitment and satisfaction tend to show no care and irresponsibility to the accomplishment of the work (low level of work performance). It is therefore crucial to regularly review employees' commitment and satisfaction to resolving any problems that may occur make sure that workers maintain a favorable attitude to work which is crucial for total organizational success.
Also, the finding supports our overarching proposition that OC will mediate the link between EE and EP. This is in line with the study conducted by Yuniarti and Prasetyaningtyas (2020), who indicated that through OC, there is a positive connection between EE and EP, and this is because dedicated workers feel positive feelings that extend their thought, allowing them to become more attentive and immersed in their job. Also, the finding supports our overarching proposition that OC will mediate the link between WE and EP. Existing empirical evidence show that WE have a positive correlativity with OC Hanaysha, 2016a;Khuong & Le Vu, 2014) and positive correlativity with EP (Badrianto & Ekhsan, 2020;Imran et al., 2012;Nguyen et al., 2015;Rorong, 2016). This implies that OC is the mechanism through which EE and WE use to increase EP. Implying that lack of EE and poor WE in one's organization and its presence in another organization serves as a push factor towards EP.
This study revealed that JS positively moderates the association between EE and OC. This is in line with previous research that found JS to positively influence EE (Chaudhry et al., 2017;Platis et al., 2015) and OC (Eslami & Gharakhani, 2012;Kirk-Brown & Van Dijk, 2016;Rusu, 2013). However, JS failed to moderate the connection between WE and OC. This study contradicts existing studies that found that JS positively affects WE (Agbozo et al., 2017;Raziq & Maulabakhsh, 2015) and OC (Eslami & Gharakhani, 2012;Kirk-Brown & Van Dijk, 2016;Rusu, 2013). The study results indicate that OC is a mediator between EE and EP and that the relationship is stronger when employee job satisfaction is high. The present finding supports what scholars highlighted, that EE and JS have positive impacts on EP (Anitha, 2014;Ayub & Islam, 2018;Chaudhry et al., 2017;Platis et al., 2015;Sendawula et al., 2018) and positive impact on OC (Eslami & Gharakhani, 2012;Hanaysha, 2016a;Imam & Shafique, 2014;Kirk-Brown & Van Dijk, 2016;Nazir & Islam, 2017;Rusu, 2013). The study results indicate that OC is a mediator between WE and EP and that the connection is stronger when employee job satisfaction is high. The present finding contradicts what scholars highlighted, that WE and JS have positive impacts on EP (Badrianto & Ekhsan, 2020;Imran et al., 2012;Nguyen et al., 2015;Rorong, 2016) and positive impact on OC Eslami & Gharakhani, 2012;Hanaysha, 2016a;Khuong & Le Vu, 2014;Kirk-Brown & Van Dijk, 2016;Rusu, 2013). The same findings were reported by Albalawi et al. (2019), who observed the variables inversely, where JS failed to moderate the link between perceived organizational support and OC. Our outcome adds to the literature's inconsistent findings; we attributed this to our approach and context; more research is needed to uncover why such association exists.

Theoretical Implications
In line with our expectations, we found that both EE and WE have a significant relationship with organizational commitment, which positively influences the performance of employees in every organization. Also, JS has a significant relationship with the performance of employees. This study will significantly contribute to the extant literature on the influence of EE, WE, and JS on OC and EP. Previous studies have equally researched into factors influencing OC and EP in different countries and firms and have recorded varying conclusions, for more reviews, see (Abdirahman, 2018;Cesário & Chambel, 2017;Eliyana & Ma'arif, 2019). However, this study will add to the literature on determinants of EP and OC in the banking sector of Ghana and other developing countries. Again, the result will add the scanty literature available and serve as a guide to other researchers about the use of sampling weights in PLS-SEM path modeling analysis.

Managerial Implications
There are several managerial implications of this study. First, our research findings advocate that managers should be familiar with the factors that influence EE and OC. This is mostly applicable in many sectors, including the service providers companies (e.g., banking) where employees are constantly facing job fatigue, which is likely to increases employee's decision to turnover. Importantly, it is prudent for managers to implement policies that will enhance OC and thus lead to EP. These policies should consider congruency between employee's JS, performance, WE, and organizational goals and aspirations. Moreover, managers should involve employees in decision-making processes to establish the factors that will increase EP and OC in their respective organizations. One significant managerial implication is seen in the finding that JS positively contributes to previous studies by impacting EP (Chaudhry, Jariko, Mushtaque, Mahesar, & Ghani, 2017;Platis, Reklitis, & Zimeras, 2015). This finding suggests that managers must consider the most effective factors that bring employees JS, which has a positive impact on the EP in an organization. Also, the study found that OC positively influences EP. Our result proposes that organizations may benefit more by promoting EP among workers dedicated to their organization.
In general terms, employees with higher levels of loyalty and fullness in an organization can mirror the positive behavior of their organizations and develop strong interests to continually work for their organizations. We suggest managers must promote employee satisfaction to win their loyalty and bring about organizational growth and development. However, employees who have less desire to be loyal to their organization may come as a result of managements' inability to provide good working conditions for employees and thus leads to poor organizational performance. It is therefore crucial for managers to maintain a good work atmosphere to encourage employees. Organizations and managers should then focus on developing the workers' workplace environment in numerous ways. This should entail valuing workers' contributions, communicating the company's progress and achievement to workers, thus instilling ownership in workers, providing them with a work-life balance, providing the requisite knowledge and tools for successful production, and providing a stable atmosphere. Management must have a place of work that guarantees the above. This will enable employees to give in their all and be committed to their organizations which serves as a driving force for organizational success and growth.

Conclusion
To a great extent, the exact estimation of people's beliefs, expectations, and values relies on how investigators choose respondents from a given population (J.-H. Cheah, Roldán, Ciavolino, Ting, & Ramayah, 2020). PLS-SEM users should be sufficiently conscious of sample collection factors, specifically when they intend their results to be relevant to the real world. Nevertheless, it is always unrealistic to obtain a sample that is reflective of a group of people. Also, the attainment of representativeness in sampling complicates problems, for instance, uneven selection probabilities, non-response, and non-coverage (Kalton & Flores-Cervantes, 2003). Investigators using PLS-SEM should participate in an ex-post modification of the sampling weights to resolve this problem and implement the WPLS-SEM algorithm of Becker and Ismail (2016), which applies the sampling weights throughout the model estimation. Our analysis of the WPLS-SEM calculations and the normal PLS-SEM evaluations makes little discrepancy in the assessment model outcomes compatible with Becker and Ismail's. However, our analysis indicates that major variations will exist if WPLS-SEM in the structural model assessment is not considered. The importance of the outcomes and the severity of the path coefficient, and the mediation effect assessment could be biased. Therefore, we propose that if investigators are interested in drawing population inferences, using the WPLS algorithm to build on less biased findings or assumptions, they should address the sampling discrepancies of their data collection. However, PLS-SEM better predicts the model in terms of Q² and the PLSpredict (RMSE, MAE, Q² predict), irrespective of both models (WPLS-SEM and PLS-SEM) having predictive relevance. All in all, our comparison reveals the weighting of the WPLS-SEM analysis of the sampling units and how they could achieve results that vary from a normal PLS-SEM analysis. This variability may have a significant effect on the empirical and managerial implications of the research and outcomes.

Limitations and Recommendations
The research did not involve people outside of Ghana. Future research may also be performed to address the limitations described by expanding the research to other settings and countries to achieve an extensive generalization of the analysis using WPLS-SEM and PLS-SEM. Future studies should also utilize the WPLS-SEM technique's effectiveness in accessing importance-performance map analysis (IPMA), multi-group analysis (MGA), and permutation analysis.

Conflicts of Interest:
The authors declare no conflict of interest.