Structural Equation Modeling Model Research Articles

Impact of Traffic Park Use on Children’s Traffic Rule Awareness and Behavioral Intentions: Case Study in Toyohashi City

To ensure children’s safe independent mobility on the road, they need to learn basic traffic rules. In this case, traffic rule education in a realistic environment through a play-way method can be a significant learning strategy. This research focuses on the “Traffic Park (TP)”, which provides that opportunity. Specifically, this research examined how elementary school children’s knowledge of traffic rules and behavioral intentions are influenced by their experiences of using and playing in a TP before and after they start school. Children from four different elementary schools (grades 1~6) were surveyed in Toyohashi City, Japan using a web-based questionnaire survey. Structural equation modeling (SEM) was applied to analyze the effect of TP use experience on children’s awareness of traffic rules and behavioral intentions. Three distinct SEM models were tested to measure the effect pattern of children’s TP use experience on their traffic rule awareness and behavioral intentions. The results show that TP use experience before entering school has a statistically significant effect on children’s traffic rule awareness (β = 0.16, p < 0.004; model 1) and behavioral intention improvement (β = 0.09, p < 0.07; model 2). However, TP use experience after entering school was found to have no significant effect. Finally, children’s TP use experience indirectly improves their behavioral intentions (β = 0.74, p < 0.001; model 3) by improving their awareness of traffic rules. Overall, the findings of this study highlight the contribution of TPs in promoting children’s safe independent mobility and fostering the development of sustainable child-friendly cities worldwide.

Exploring User Experience in Virtual Industrial Heritage Platforms: Impact of Cultural Identity, Functional Clarity, Scene Interactivity, and Narrative Quality

This study aims to investigate the key factors influencing user experience in the design of virtual interactive platforms for the preservation of industrial heritage within the context of digitalization. In the literature review, this paper introduces a theoretical model comprising five latent variables: cultural identity, functional clarity, scenario interactivity, narrative quality, and user experience. To construct and validate the model, the author employed structural equation modeling (SEM) to analyze data from 323 valid questionnaires collected in China. The validation of the SEM model led to conclusions regarding the impact of each latent variable on user experience. The results indicate that cultural identity, functional clarity, scene interactivity, and narrative quality significantly affect user experience and play a critical role in enhancing user experience on virtual platforms. To validate this theoretical framework, the study employs the virtual interactive platform of Hanyang Ironworks as a case study, examining how these factors manifest in practical scenarios and their specific impact on platform design. The findings reveal that a strong sense of cultural identity, clear functional design, interactive scenes, and a well-structured narrative are the core factors that enhance user experience on virtual industrial heritage platforms. This research provides both theoretical support for the digital preservation of industrial heritage and practical insights for the design of virtual platforms. The study improves virtual interactions with industrial heritage and informs future research and applications.

Analysis of interrelationships among constructs of cost overrun causes in consortium projects between construction contractors

Cost overruns can negatively affect consortium contractors, leading to financial difficulties, reduced profits, reputation damage, and even bankruptcy. Thus, it is crucial to identify and analyze the causes of cost overruns in consortium projects between contractors. However, there has been limited research on consortium projects between contractors. This study fills a critical research gap by specifically investigating the causes of cost overruns and their interrelationships in consortium projects between contractors, a domain that is globally prevalent but has a limited empirical focus. Questionnaire surveys are used to collect data from practitioners. The study identifies thirty-three causes of cost overruns. Using Structural Equation Modeling (SEM), the study uncovers the interrelationships among eight constructs of cost overrun causes, including construction-related problems, consortium contract-related problems, management-related problems, finance-related problems, culture-related problems, dispute-related problems, legal-related problems, and external objective problems. This study provides a deep understanding of the interrelated constructs by focusing on the results of the SEM model. This not only mitigates the risk of cost overruns but also provides practical guidelines for drafting robust contracts and managing consortium projects between contractors effectively.

On Selecting Indicators for SEM Models: Trade-Offs Among Variable Type, Quantity, and Quality

To study the dynamics among latent variables, researchers often use structural equation modeling (SEM). The real interest of SEM is often in the structural part, which explains the relationships among the latent variables. The indicators are peripheral; they exist because it is impossible to obtain the latent variables’ scores. Perhaps because the indicators are peripheral in developing substantive theories, how to choose indicators for SEM has received little attention in the literature. The attention has been largely drawn to model estimation and evaluation given a set of indicators, but the importance of indicators for the quality of model estimation has been overlooked. In this paper, we study the trade-offs among the variable types, quantity, and quality of indicators with respect to the quality of estimating structural-level model parameters. Better selections of indicators can substantially improve estimation quality without increasing the sample size. Using many lower-loading indicators is not necessarily inadvisable.

Disaggregated effects of artificial intelligence, online and mobile banking on customer satisfaction in banks: An analysis using structural equation modelling

In the Fourth Industrial Revolution (4IR) era, the rapid digitalisation of services poses both opportunities and challenges for the banking sector. This study addresses how adopting artificial intelligence (AI) and online and mobile banking advancements can influence customer satisfaction, particularly in Kaduna State, Nigeria. Despite significant investments in AI and digital banking technologies, banks often struggle to align these innovations with customer expectations and satisfaction. Using Structural Equation Modeling (SEM), this research investigates the impact of customer satisfaction with online banking (C_O) on AI integration (I_A) and mobile banking convenience (C_M). The SEM model reveals that customer satisfaction with online banking significantly influences AI integration (path coefficient of 0.40) and mobile banking convenience (path coefficient of 0.68). These results highlight a crucial problem: while technological advancements in banking are growing, their effectiveness is highly dependent on customer satisfaction with existing digital services. The study underscores the need for banks to prioritise enhancing online banking experiences as a strategic lever to improve AI integration and mobile banking convenience. Consequently, the research recommends that Nigerian banks develop comprehensive frameworks to evaluate and optimise their technology integration strategies, ensuring that technological innovations align with customer needs and expectations in the rapidly evolving digital landscape.

The Influences of Corporate Governance on Corporate Social Responsibility and Firm Performance of the Listed Companies in Thailand Sustainability Investment

This study investigates the impact of corporate governance and social responsibility on the financial performance of 149 sustainable stocks listed on the Stock Exchange of Thailand using secondary data from 56-1 reports and employing structural equation modeling (SEM) to assess model fit, the analysis identified several governance factors, such as the proportion of independent directors, board size, and board meeting frequency, as significantly influencing market value. The SEM model demonstrated a strong fit with empirical data (CMIN/df=1.504, CFI=0.990, RMSEA=0.058), with board meeting frequency showing the most significant impact on financial performance at the 0.05 level. The effective corporate governance, particularly in terms of board structure and oversight, is crucial for enhancing the market value of sustainability-focused firms.

MANAGING WORK-LIFE BALANCE TO OPTIMIZE EMPLOYEE WELFARE AND PRODUCTIVITY

This study investigates the impact of work-life balance on employee well-being and productivity within the public sector, specifically focusing on the East Luwu Regency. Utilizing Structural Equation Modeling (SEM), the research examines how work-life balance contributes to employee welfare and overall productivity. The study draws on data collected from 200 employees across various departments within the local government. The results reveal that work-life balance significantly and positively influences both employee well-being and productivity, with the SEM model indicating that 51.5% of the variance in well-being and 55.0% in productivity can be explained by work-life balance. These findings underscore the critical role of work-life balance in enhancing employee outcomes and highlight the need for local governments to adopt more flexible and supportive work policies. The research contributes to the broader understanding of organizational behavior within the public sector and offers practical recommendations for improving employee welfare and productivity through better work-life balance strategies.

DETERMINANTS IN MODELLING EARLY SCHOOL DROPOUT

Recognizing the importance of continuity in education, it was deemed necessary to carry out a study whose main objective is to identify the factors that lead to the intention to drop out in order to formulate effective strategies to combat this phenomenon. Early school dropout has negative consequences both for individual development and for social and economic progress. Therefore, this paper aims to contribute to the understanding and prevention of this phenomenon through a rigorous analysis of its determinants. In this context, eight relevant dependent variables have been identified in the literature that are believed to play a significant role in the intention to drop out of school. These variables include factors such as school absenteeism, alcohol or substance abuse, attitude, awareness, family, family supervision, school environment and school rules. The analysis used in the study examines these significant variables through structural equation modeling (SEM). SmartPLS software was used to conduct this analysis, which allows the use of Partial Least Squares SEM (PLS-SEM) and Bootstrapping modeling techniques. The data used for this research was collected using a well-structured questionnaire consisting of 28 questions aimed at capturing students' perceptions and experiences of school and the factors that might contribute to their intention to drop out of school. A total of 669 respondents completed the questionnaire, providing a solid database for analysis.

Mediation effect of neuropsychological factors on the relationship between social networks and mild cognitive impairment in older adults.

Although links between social networks and mild cognitive impairment (MCI) have been suggested. The pathway between social networks and MCI from the Person-Environment-Occupation (PEO) model perspective among older adults remains inconclusive. To examine the relationship between social networks and MCI, and further identify whether neuropsychological factors mediate the relationship. 1036 participants aged 55 to 75 years were recruited from two districts of Fuzhou, China, from December 2020 to December 2022. Data were obtained via a face-to-face neuropsychological scale assessment. Social networks and cognitive function were assessed by the Lubben Social Network Scale and the Montreal Cognitive Assessment, respectively. The mediation model and structural equation model (SEM) pathway analysis were used to examine the direct and indirect effects of social networks on MCI via neuropsychological factors. Data from 580 participants were analyzed (year: 65.16 ± 5.38). Regression analysis indicated that higher levels of social networks and psychological resilience were positively correlated with improved cognitive function, even after adjusting for demographic data. Increased depressive symptoms and poor sleep quality were linked to cognitive decline. Support from family members had a greater impact than support from friends in reducing the risk of developing MCI. The SEM model supported the hypothesis that significant indirect effects of social networks on MCI via psychological resilience, depressive tendencies, and sleep quality. The effects of social networks on MCI are mediated by psychological resilience, depressive tendencies, and sleep quality.

Promoting Sustainable Safety Work Environments: Factors Affecting Korean Workers’ Recognition of Their Right to Refuse Dangerous Work

(1) Background: The right to refuse dangerous work (RTRDW) is essential for preventing industrial accidents and protecting worker safety in Korea. However, its use remains limited in practice. This study seeks to identify the factors hindering its activation across industries such as construction, manufacturing, and services, offering a comprehensive analysis beyond previous research. (2) Methods: A survey was conducted across key industries to assess five factors—safety behavior, communication, management commitment, education and training, and education and training—using structural equation modeling (SEM) to evaluate their influence on the exercise of RTRDW. (3) Results: The SEM model showed a good fit (χ2 = 1151.333, p < 0.001, TLI = 0.978, CFI = 0.984, RMSEA = 0.05). The most significant factors influencing RTRDW were safety performance behavior and communication, while ambiguous regulations, poor training, and fear of job loss discouraged its use. (4) Conclusions: To improve RTRDW activation, clearer regulations, enhanced safety education and training, stronger management commitment, and better communication are necessary. Addressing these issues can help workers confidently exercise their right to refuse dangerous work, enhancing overall workplace safety. (5) Benefits: This study provides practical strategies for policymakers and industry leaders to promote safety, empowering workers to use RTRDW effectively and contributing to a safer work environment.

Attitudes and Motivation as Predictors of Perception of English Language Proficiency: Structural Equation Modeling (SEM)

English is the most widely spoken and written language in the world, so many students have different perceptions of English language proficiency, so there is a need to investigate the predictors of English language learning. Therefore, the main objective of this study was to determine the relationship between attitudes towards learning English and motivation towards English learning under the SEM model as predictors of perception of English language proficiency in Peruvian university students. The sample consisted of 711 Peruvian students from a private university (442 females = 62.2% and 269 males = 37.8%). The questionnaire of attitudes towards learning English, motivation towards English, and socio-demographic questions were used, including the perception of English language proficiency. The proposed model obtained a satisfactory fit. Attitudes towards learning English explain 10% of the perception of English language proficiency, but with motivation towards English as the mediating variable. According to the established criteria, the goodness of fit of the model was confirmed (?2 = 383.78; df =3; p .000; RMSEA = 0.06; AGFI= 0.97; TLI = .97; CFI = .99), which confirms the fact that all the values obtained are acceptable. Received: 10 May 2024 / Accepted: xx October 2024 / Published: 05 November 2024

Influencing factors of kinesiophobia in knee arthroplasty patients under the social cognitive theory: A structural equation model

ObjectiveTo analyze the path relationships among influencing factors for kinesiophobia in knee arthroplasty patients through a structural equation model. BackgroundThe occurrence of kinesiophobia significantly impacts the rehabilitation process of knee arthroplasty patients. However, there is still a need to determine factors that contribute to reducing kinesiophobia. DesignA cross-sectional study was conducted and reported following the STROBE guideline. MethodsBetween February 2022 to October 2022, 162 total knee arthroplasty (TKA) patients and 81 unicompartmental knee arthroplasty (UKA) patients completed a survey. A structural equation modeling (SEM) approach was utilized to analyze the relationships between kinesiophobia and influencing factors (social support, pain resilience, and rehabilitation self-efficacy). Furthermore, multi-group SEM analysis was conducted to examine whether the model equally fitted patients in different types of knee arthroplasty. ResultsThe direct negative effects of rehabilitation self-efficacy (β = −0.535) and pain resilience (β = −0.293) on kinesiophobia were observed. The mediating effect (β = −0.183) of pain resilience and rehabilitation self-efficacy between social support and kinesiophobia was also significant. The SEM model achieved an acceptable model fit (χ2 = 35.656, RMSEA = 0.031, χ2/df = 1.230, GFI = 0.972, NFI = 0.982, IFI = 0.997, CFI = 0.996). In multicohort analysis, no significant differences were observed among knee arthroplasties (TKA, UKA) (Δχ2 = 4.213, p = 0.648). ConclusionsSatisfactory social support enhances pain resilience and rehabilitation self-efficacy, so as to reduce kinesiophobia. Future interventions that directly target the assessment and management of kinesiophobia, available social support may help reduce kinesiophobia, and pain resilience and rehabilitation self-efficacy may be critical factors in managing kinesiophobia. Relevance to Clinical PracticeReducing kinesiophobia in knee arthroplasty patients requires satisfactory social support, pain resilience, and rehabilitation self-efficacy. Therefore, healthcare organizations may implement initiatives to reduce kinesiophobia by taking these factors into account.

Decoding the impact of political uncertainty on IPO underpricing in China

PurposePolitical factors play a crucial role in China's initial public offering (IPO) market due to its distinctive institutional context (i.e. “economic decentralization” and “political centralization”). Given the significant level of IPO underpricing in China, we examine the impact of local political uncertainty (measured by prefecture-level city official turnover rate) on IPO underpricing.Design/methodology/approachUsing 2,259 IPOs of A-share listed companies from 2001 to 2019, we employ a structural equation model (SEM) to examine the channel (voluntarily lower the issuance price vs aftermarket trading) through which political uncertainty affects IPO underpricing. We check the robustness of the results using bootstrap tests, adopting alternative proxies for political uncertainty and IPO underpricing and employing subsample analysis.FindingsLocal official turnover-induced political uncertainty increases IPO underpricing by IPO firms voluntarily reducing the issuance price rather than by affecting investor sentiment in aftermarket trading. These relations are stronger in firms with pre-IPO political connections. The effect of political uncertainty on IPO underpricing is also contingent upon the industry and the growth phase of an IPO firm, more pronounced in politically sensitive industries and firms listed on the growth enterprise market board.Originality/valueLocal government officials in China usually have a short tenure and Chinese firms witness significantly severe IPO underpricing. By introducing the SEM model in studying China IPO underpricing, this study identifies the channel through which local government official turnover to political uncertainty on IPO underpricing.

Healthcare Access Domains and Treatment as Mediators of Ovarian Cancer Racial Disparities in Survival: A Structural Equation Modeling Analysis in SEER-Medicare.

Racial differences in healthcare access (HCA) may contribute to disparities in ovarian cancer (OC) survival. We used structural equation models (SEM) to examine associations between race and HCA domains (affordability, availability, accessibility) in relation to overall and OC-specific mortality. Non-Hispanic (NH)-Black and non-Black (Hispanic, NH-White) women diagnosed with OC in 2008-2015 were identified from SEER-Medicare. Cox proportional hazards regression was used to conduct mediation analysis for associations between race and HCA domains with overall and OC-specific mortality. SEM models adjusting for demographic and clinical covariates were used to estimate hazard ratios (HR) and 95% confidence intervals (CI). A total of 4,629 eligible OC patients were identified, including 255 (5.5%) patients who were NH-Black. In SEM adjusting for demographic, clinical, and HCA latent variables, there was a total effect of NH-Black race on overall (HR: 1.11, 95% CI: 1.03,1.19) and OC-specific mortality (HR: 1.16, 95% CI: 1.08, 1.24), which was primarily driven by a direct effect. There was a modest indirect association between NH-Black race and mortality through decreased treatment receipt, though not through HCA. There is a need for studies investigating additional social and biological mechanisms that contribute to worse cancer survival among NH-Black patients.

Safety-risk assessment system for prefabricated building construction in China

PurposePrefabricated building (PB) uses factory production and onsite assembly, which differs from traditional construction methods. This special construction approach may lead to dissimilar safety risks and challenges. Traditional safety assessment methods may not adequately and accurately assess the safety risks of PB construction. This paper aims to develop a new concept and methodology for targeted improvement in assessing PB safety risks.Design/methodology/approachRisk factors and indicators were established based on literature review and expert inputs. A structural equation model (SEM) was developed to investigate the relationships among three main risk categories: objects, workers and management. SEM analyzed the intricate associations between indicators and deepened understanding of safety risks. The model was tailored for China’s PB construction projects to enhance safety-risk management.FindingsThe cloud model evaluation validated the SEM model. A PB case study project tested and verified the model, evaluated its efficacy and quantified its safety performance and grade. We identified significant safety risk impacts across the three risk categories, safety-control level and specific areas that require improvement. The SEM model established a robust safety evaluation indicator system for comprehensive safety assessment of PB construction.Practical implicationsPractical recommendations provide valuable insights for decision-makers to enhance construction efficiency without compromising safety. This study contributed to the conceptual foundation and devised a novel method for evaluating safety performance in PB construction for safer and more efficient practices.Originality/valueThis study departed from the traditional method of calculating weights, opting instead for the SEM method to determine the weights of individual risk indicators. Additionally, we leveraged the cloud model to mitigate the influence of subjective factors in analyzing questionnaire survey responses. The feasibility and reliability of our proposed method were rigorously tested and verified by applying it to the PB case.

Modelling the factors associated with quality of life in women with osteoporosis: A cross-sectional study

BackgroundConsidering the important factors contributing to different health-related quality of life (HRQoL) subscales is essential for implementing preventive measures to increase the HRQoL among women with osteoporosis. We here evaluated the variables related to the mental and physical HRQoL in a sample of Iranian osteoporotic women. MethodsIn this cross-sectional study, the participants included 111 women with osteoporosis in 2013. Physical and mental of HRQoL were measured by the ECOS-16 questionnaire. Other variables included BMD t-score (Osteoporosis was diagnosed based on bone mineral density (BMD) with BMD t-score < −2.5), age, body mass index, educational level, marital status, duration of the disease, history of bone fracture in the past year, the number of pregnancies, and weekly walking hours. Final regression coefficients were obtained based on the total effects of estimations (decompositions of effects into direct, indirect and total effects) by structural equation model (SEM) analysis. ResultsThe mean scores of physical and mental domains of HRQoL were 54(21) and 54(25), respectively. The mean of BMD t-score was −3.2 (0.9). Based on the regression coefficients obtained in the SEM model, weekly walking hours(2.2), number of pregnancies (−1.2), and history of bone fracture in past year(−2.9) were the more important determinants of physical aspect of HRQoL than other included variables of this study. Furthermore, age over than 60 (−9.1), history of bone fracture in past year(−4.8), weekly walking hours(2.3), marital status(−5), and education level (3.9)influenced the mental aspect of HRQoL. ConclusionsSocial and life style factors tend to impact on physical and mental domains of HRQoL, a measure that is influenced by multiple factors among postmenopausal women. In this respect, these obtained factors should be considered for health planning to improve the physical and mental domains of HRQoL among postmenopausal women.

Innovation capability as a mediator of the relation between talent management and entrepreneurial firm performance

PurposeThe study aims to investigate innovation capability as a mediator in the relationship between talent management (TM) and entrepreneurial firm performance in Micro, Small and Medium Enterprises (MSMEs) in Thailand based on employer perceptions.Design/methodology/approachA questionnaire survey was completed by 674 MSMEs in Thailand, both directly and online. Four hypotheses were tested using structural equation modelling (SEM).FindingsThe results demonstrated that TM has a significant and positive influence on innovation capability, but no evidence was found for a positive relationship between TM and entrepreneurial firm performance. However, innovation capability positively influences entrepreneurial firm performance, and it fully mediates the positive relationship between TM and entrepreneurial firm performance. These results suggest that TM can enhance entrepreneurial firm performance if it is implemented with effective innovation capability.Research limitations/implicationsThe SEM model was employed with a limited number of enterprises, so the sample cannot represent the full population, as there are more than three million MSMEs in Thailand. More countries and factors should be incorporated into the model.Practical implicationsThe owner-managers and HR practitioners of MSMEs can utilise the study results to develop talent strategies and innovative interventions that will enhance their firm’s performance.Social implicationsThis study thus aims to strengthen the capacities of MSME entrepreneurs and policymakers to elevate inclusive innovation capability in MSMEs and to also accelerate progress towards the Sustainable Development Goals (SDG); innovation is one of the three important aspects of SDG9.Originality/valueThe study contributes by filling the research gap related to TM in the MSME business and enterprise development literature, and it aids in creating innovation capability to increase effective investments in TM.

The Influence of Vegetation Environment on Thermal Experience in Hot Summer: A Case Study from Perspectives of Fitting Scale and Gender Disparity

Vegetation exerts a significant cooling effect, particularly during the hot summer; however, the spatial scale effects and gender difference among occupants’ subjective thermal comfort remain elusive. Developing a comprehensive model to elucidate the multidimensional relationship between green spaces and thermal experience holds paramount importance. Taking Longzi River Park in Zhengzhou city as a case study, this research examined the influence of vegetation on thermal experience by using structural equation modeling (SEM) from perspectives of fitting scale and gender disparities. It was found that (1) The vegetation environment not only influences thermal sensation, comfort and demand independently, but also influences the pathway between them. These influence paths constitute a complex causal network, functioning as a framework of “sensation → comfort → demand” and its influencing factors. (2) There exists a scaling effect in the pathway framework, which conforms to a threshold of 10 m for the inner radius and 30 m for the outer radius. The goodness of SEM model fit declines with the increase in either the inner radius or the outer radius, or both. (3) Differences in genders are exhibited for the pathway framework, with the vegetation exerting a stronger influence on female sensation and comfort, as well as male demand. The pathway from sensation to comfort to demand is more pronounced in male populations. The research findings contribute to the development of improved and sustainable vegetation distribution in urban parks.

Understanding the relationship between perceived accessibility, housing and transport equity of different types of residents: A structural equation modelling approach

This study uses structural equation modelling (SEM) to analyse the relationship between perceived accessibility, housing and transport equity and their influencing factors in Nanjing, China. It aims to address the current shortfall in studies to compare differences in perceived (rather than objective) measures of accessibility using a transport equity lens. A comparison sample of 664 affordable housing residents and 828 commercial housing residents living in different locations across the city was collected through an online survey. The research shows that affordable housing residents are in a more disadvantaged position than commercial housing residents based on their perceived accessibility, housing and transport equity evaluation results are all worse than those of commercial housing residents. Therefore, two SEM models were constructed to understand the underlying factors driving these perceptions. The results show that both housing location and transport provision significantly affect perceived accessibility, housing and transport equity. Improving people’s transport affordability is the most effective way to improve perceived accessibility, and transport equity for both groups, however, improving the housing environment is more critical for affordable housing residents, while improving housing affordability is more important for commercial housing residents.

Factors affecting employee retention: An empirical study in Nepalese commercial banks

This paper aims to identify the factors affecting employee retention in Nepalese financial institutions, especially commercial banks. Through a questionnaire, the primary data were collected from 400 respondents. The study is based on a dichotomous and 5-point Likert scale questionnaire. The collected data were analyzed using SPSS and structural equation modeling (SEM). SEM smart PLS-4 is used to examine the association between the dependent and independent variables. The results indicate that company image/goodwill, employee benefits, job satisfaction, organizational performance, recruitment and selection, and work environment and culture have a relatively significant relationship with employee retention. Among them, organizational performance was found to be highly significant among all other variables. However, labor unions, leadership support, performance appraisal, training and development, and work-life balance did not have an association with employee retention. The study helps HR managers to formulate appropriate policies and guidelines to retain employees in Nepalese commercial banks. Likewise, it provides a valuable empirical contribution by analyzing which factors affect employee retention in commercial banks in Nepal using a direct SEM model. It also shows which factors play a major role in employee retention.