Work-life balance is a crucial issue in modern human resource management, particularly in the banking sector, which demands high performance and long working hours. An imbalance between work and personal life can trigger stress, reduce motivation and increase turnover employee intentions [1]. This condition requires companies to create a healthier work environment and support employee well-being.

Furthermore, job satisfaction plays a significant role in reducing turnover intention. Robbins and Judge stated that employees who are satisfied with their jobs tend to have higher loyalty to the organization. However, there is still a phenomenon of employees in the banking sector choosing to leave their companies despite receiving relatively competitive compensation, indicating other factors influencing turnover intention.

Work engagement is also a crucial variable in understanding employee behavior. Schaufeli et al. [2] explain that employees with high levels of work engagement demonstrate dedication, vigor and absorption, which can improve performance while reducing the desire to leave. However, practical experience shows that engagement is not entirely effective in suppressing turnover intentions.

In this context, mentoring is seen as a factor that can strengthen the influence of these variables on turnover intention. According to Kram, mentoring plays a role in providing career and psychosocial support, which ultimately increases employee job satisfaction and engagement. Therefore, this study positions mentoring as a moderating variable that has the potential to weaken the negative relationship between work-life balance, job satisfaction and work engagement on turnover intention.

Based on the above phenomena, the formulation of the research problem is: (1) Does work-life balance affect turnover intention? (2) Does job satisfaction affect turnover intention? (3) Does work engagement affect turn-over intention? and (4) Is mentoring able to moderate the influence of work-life balance, job satisfaction and work engagement on turnover intention?

While much previous research has examined the direct relationship between these variables, gaps remain. For example, Karatepe's [3] study found that work-life balance significantly influences turnover intention, while Haar et al.'s study yielded conflicting results. These discrepancies in findings highlight inconsistencies that require further investigation.

Furthermore, most previous research has focused solely on work-life balance, job satisfaction or work engagement separately, without linking them to the moderating role of mentoring. However, mentoring is believed to significantly contribute to strengthening the relationship between variables, particularly in the context of service organizations that require intensive interaction, such as banking.

The purpose of this study is to analyze the influence of work-life balance, job satisfaction and work engagement on turnover intention, as well as to examine the role of mentoring as a moderating variable. Thus, this study is expected to provide a more comprehensive understanding of the factors influencing turnover intention in service organizations.

This research contributes to addressing the research gap related to inconsistencies in previous findings and the lack of studies that consider mentoring as a moderating variable. Theoretically, this research enriches the literature on organizational behavior and human resource management. Practically, this study provides recommendations for bank management or service companies to formulate strategies to improve employee welfare, job satisfaction, engagement and effective mentoring programs to reduce turnover intention.

Thus, this research not only provides academic value, but also practical implications for organizations in creating a sustainable and conducive work climate, especially in the service sector which has a high level of competition.

Literature Review and Hypothesis Development

In an effort to understand the factors that influence turnover intention, various previous studies have examined the role of work-life balance, job satisfaction and work engagement as the main variables.

All three have been shown to influence employee loyalty and engagement with the organization. Furthermore, mentoring is also gaining attention as a mechanism for strengthening these relationships.

Work-life balance is a crucial issue in human resource management, relating to an individual's ability to balance the demands of work and personal life. Employees with a good work-life balance tend to have higher motivation and loyalty, as well as lower stress levels [1]. According to Delecta, this balance encompasses work obligations, family responsibilities and other social commitments.

The dimensions of work-life balance can be seen through the interaction between work and personal life. Fisher et al. in Gunawan explain two main dimensions, while Hayman adds three dimensions: Work Interference with Personal Life (WIPL), Personal Life Interference with Work (PLIW) and Work/Personal Enhancement Life (WPEL). These dimensions emphasize that balance can have both positive and negative impacts on turnover intention.

Job satisfaction is also a crucial factor that reflects the psychological well-being of employees. Robbins and Judge emphasize that job satisfaction encompasses compensation, quality of work relationships, career opportunities and company policies. Satisfied employees tend to have higher levels of commitment, while dissatisfaction often leads to increased turnover intentions.

Job satisfaction is viewed as a subjective experience, where each individual evaluates their work according to their preferences and expectations. Dimensions of job satisfaction include cognitive, affective and behavioral aspects. According to Spector, the behavioral dimension is more informative because it provides a concrete indication of a person's level of job satisfaction, for example through attendance, work enthusiasm and desire to stay.

Work engagement, according to Schaufeli et al., consists of vigor, dedication and absorption, which describe an employee's attachment to their work. Employees with high levels of engagement demonstrate strong motivation and commitment, resulting in lower turnover intentions. However, other research has also found that engagement can be influenced by external factors such as organizational culture and management policies [4].

Mentoring is seen as a crucial variable that can strengthen the relationship between work-life balance, job satisfaction, work engagement and turnover intention. Kram stated that mentoring functions in two aspects: career support and psychosocial support. Ragins and Kram added that effective mentoring can increase job satisfaction, strengthen engagement and reduce employee turnover intentions.

Several empirical studies have shown mixed results. Karatepe [3] found that work-life balance significantly influences turnover intention, while Haar et al. reported inconsistent results. Similarly, the relationship between job satisfaction and turnover intention is sometimes attenuated by external factors, such as career opportunities outside the organization [5]. This indicates a research gap that still needs to be filled.

Based on the theory and results of previous research, the following hypotheses can be formulated: (H1) Work-life balance has a negative effect on turnover intention. (H2) Job satisfaction has a negative effect on turnover intention. (H3) Work engagement has a negative effect on turnover intention. (H4) Mentoring moderates the effect of work-life balance on turnover intention. (H5) Mentoring moderates the effect of job satisfaction on turnover intention. (H6) Mentoring moderates the effect of work engagement on turnover intention.

Thus, the hypothesis development in this study stems from key theories on work-life balance, job satisfaction, work engagement and mentoring and is supported by previous empirical findings. This strengthens the conceptual basis that these variables interact to influence turnover intention, particularly in the context of service organizations such as banking.

This study aims to analyze the factors influencing turnover intention among bank employees, focusing on independent variables, namely work-life balance, job satisfaction and work engagement, with mentoring as a moderating variable. Turnover intention is defined as an employee's voluntary intention to leave an organization, which can be influenced by factors such as job satisfaction, job involvement and work-life balance [5,6]. In the context of the banking industry, which has a high workload and tight targets, this issue is important to study [7].

This study employed a quantitative approach with an explanatory survey method. An explanatory design was chosen because it is able to explain causal relationships between research variables objectively and measurably. The primary research instrument was a closed-ended questionnaire constructed using a Likert scale, allowing for numerical measurement of respondents' perceptions of the research variables.

The research population is bank employees who are part of the millennial generation, namely individuals born between 1981 and 2000. Strauss and Howe, who work at head offices and branches in the Greater Jakarta area. This generation was chosen because it has a tendency towards high turnover intention due to different work expectations and emotional attachments compared to previous generations.

The sample size was determined using the 10-times rule approach proposed by Hair et al., which requires a minimum of 10 times the number of paths in the research model leading to the dependent variable. Because this study has four main paths, the minimum sample size required is 40 respondents. The sampling technique used was proportional stratified random sampling to ensure proportional representation of each division and branch.

The data used in this study were primary data obtained through a Google Form-based questionnaire sent to the company's official email address. This technique was chosen because it allowed respondents to provide written answers systematically and more efficiently.

The questionnaire uses a six-point Likert scale, ranging from "strongly disagree" (score 1) to "strongly agree" (score 6), so that it is able to capture variations in respondents' perceptions in more detail.

Prior to the main analysis, instrument validity and reliability were tested. Validity was tested using Pearson correlation with criteria of r>0.3 and significant at p<0.05. Reliability was tested using Cronbach's Alpha, with a value of ÿ 0.70 considered reliable. This is crucial to ensure the research instrument can consistently measure the constructs.

The next stage is testing the classical assumptions, namely normality, multicollinearity and heteroscedasticity.

The normality test uses Kolmogorov-Smirnov or Shapiro-Wilk, the multicollinearity test uses the Tolerance value (<0.1) and VIF (<10), while heteroscedasticity is tested using the Glejser test or residual scatterplot pattern.

To analyze the direct influence between variables, multiple linear regression was used with the equation model:

(1)

Where;

TI                            = turnover intention

WLB       = work-life balance

KK                           = job satisfaction

WE                         = work engagement

ÿ                              = error

Next, the moderating role of mentoring was tested using moderated regression analysis (MRA) by forming an interaction variable (e.g., WLB × Mentoring).

Data analysis was conducted using Structural Equation Modeling (SEM) based on covariance-based SEM (CB-SEM). SEM was chosen because it is able to test complex theoretical models, involving direct and indirect relationships and includes moderating variables in one analytical framework. The measurement model was tested using convergent validity (loading factor>0.7 and AVE>0.5), discriminant validity (the square root of AVE is greater than the correlation between constructs) and reliability through Cronbach's Alpha and Composite Reliability>0.7.

The structural model is evaluated through the path coefficient, t-statistic value (<1.96 at ÿ = 0.05) and R² value which shows the proportion of variance in the dependent variable explained by the independent variable.

Mentoring moderation is said to be significant if the interaction coefficient is significant and the R² value increases after the interaction variable is entered.

This research was conducted at a state-owned bank focused on housing finance, specifically its branch in the Semarang area. This bank plays a strategic role in supporting government programs in the housing sector and is a pioneer in distributing subsidized and non-subsidized home ownership loans (KPR). With a vision of becoming the Best Mortgage Bank in Southeast Asia by 2025, the company prioritizes employee welfare, digitalization and an AKHLAK work culture as the main foundations of its operations.

Before conducting hypothesis testing, a classical assumption test was carried out to ensure the feasibility of the regression model.

The normality test using Kolmogorov-Smirnov produced a significance value of 0.071 (<0.05), so that the residuals were declared to be normally distributed (Table 1).

Table 1: Normality Test

*Test distribution is normal

The heteroscedasticity test using the Glejser method shows that all independent variables have a significance value above 0.05, so the model is free from heteroscedasticity (Table 2).

Table 2: Heteroscedasticity Test Coefficients

Dependent Variable: ABS_RES

Furthermore, the multicollinearity test showed a tolerance value>0.10 and VIF<10, indicating that there was no multicollinearity between the independent variables (Table 3).

Table 3: Multicollinearity Test Coefficients

Dependent Variables: Turnover Intention (Y)

The results of multiple linear regression indicate that work-life balance (WLB), job satisfaction (KK) and work engagement (WE) influence turnover intention (TI). The regression equation obtained is: TI = 8.498+ 0.028WLB + 0.286KK + 0.314WE + e. The constant value of 8.498 indicates that turnover intention remains high even though the three independent variables are at constant values. The regression coefficient shows that work-life balance has a positive but insignificant effect, while job satisfaction and work engagement have a greater influence on turnover intention (Table 4).

Table 4: Multiple Linear Regression Analysis Coefficients

Dependent Variable: Turnover Intention (Y)

A partial t-test showed that the work-life balance variable was not significant on turnover intention (sig. 0.873>0.05). This result differs from the findings of Haar et al. who stated that work-life balance can reduce turnover intention, but supports Byron's view that the impact of work-life balance is often influenced by contextual factors such as organizational culture and job characteristics.

Thus, work- life balance is not a primary determinant of employee turnover intention in the context of this study. Job satisfaction showed a positive effect on turnover intention (sig. 0.068>0.05), but was not significant at the 5% level. This result aligns with research by Robbins and Judge, which states that job satisfaction is closely related to loyalty, but is inconsistent with Locke, who emphasized job satisfaction as the primary predictor of retention. This indicates that compensation or intrinsic satisfaction alone is not sufficient to prevent turnover intention, given that work demands in the banking sector often exceed the incentives provided.

Work engagement has been shown to have a significant effect on turnover intention, with a significance value of 0.006 (<0.05). This confirms that employees with high vigor, dedication and absorption tend to be more committed to the organization, making them less likely to leave their jobs [2]. This finding is consistent with work engagement theory, which positions engagement as a key factor in reducing turnover intentions.

The results of the moderation test indicate that mentoring strengthens the relationship between job satisfaction and work engagement and turnover intention, although it is not significant for work-life balance. This finding supports research by Allen and Meyer [8] and Kram, which states that mentoring can provide emotional and career support, so employees feel valued and more motivated to stay. Mentoring is more effective in increasing engagement and satisfaction than simply improving work- life balance.

Overall, the study results answer the research question: work engagement is a significant factor in reducing turnover intention, while work-life balance and job satisfaction show weaker effects. The role of mentoring as a moderating variable underscores the importance of managerial intervention in employee retention, rather than relying solely on individual factors such as satisfaction or work-life balance.

The findings of this study contribute to explaining inconsistencies in previous studies. For example, Karatepe [3] emphasized the strong influence of work-life balance on turnover intention, but in this study, the effect was insignificant. By including mentoring as a moderating variable, this study demonstrates the important role of organizational contextual factors in strengthening or weakening the relationship between variables.

The practical implication of this research is the need for organizations, particularly in the banking sector, to focus more on improving employee engagement through systematic mentoring programs. This aligns with Ragins and Kram's recommendation, which emphasizes the importance of mentor support in building loyalty. Thus, employee retention strategies can be directed not only at increasing compensation but also at strengthening emotional attachment and supporting career development.

Based on the results of research conducted on employees of a state-owned bank regarding the influence of work-life balance, job satisfaction and work engagement on turnover intention with mentoring as a moderating variable, the following conclusions can be drawn:

• The constant value of 8.498 in the regression model indicates that when all independent variables, namely work-life balance (WLB), job satisfaction (KK) and work engagement (WE), are at zero or considered to have no effect, then turnover intention (TI) remains at 8.498. This reflects the existence of a basic level of turnover intention that is not influenced by these three variables, so there is a possibility that there are other external factors outside this research model that also influence employees' desire to leave the company

• The results of the Moderated Regression Analysis (MRA) test show a significant interaction value between work-life balance and mentoring (WLB_MI) of 0.340 (<0.05). This means that mentoring does not significantly moderate the effect of work-life balance on employee turnover intention. Although mentoring generally influences turnover intention (p-value 0.009), its presence has not been able to strengthen or weaken the effect of work-life balance on turnover intention. This indicates that the implemented mentoring program is not yet effective enough to optimize the role of work-life balance in reducing turnover intention

• The results of the moderation analysis also show that the significance value of the interaction between mentoring and job satisfaction (KK_M) is 0.176 and between mentoring and work engagement (WE_MI) is 0.551, both of which are greater than 0.05. Thus, mentoring does not significantly moderate the effect of job satisfaction or work engagement on turnover intention. This indicates that although job satisfaction and work engagement have been shown to influence turnover intention, the presence of mentoring does not significantly change the strength of the relationship. This condition confirms that the effectiveness of mentoring programs in increasing employee retention is still not optimal, so evaluation and development are needed so that mentoring can become a stronger strategy in strengthening the relationship between internal factors of employees and their intention to stay in the organization

1. Greenhaus, J.H. et al. “The relationship between work-family balance and quality of life.” Journal of Vocational Behavior, vol. 63, no. 3, 2003, pp. 510–531.

2. Schaufeli, W.B. et al. “Work engagement, job demands and job resources as predictors of absenteeism.” 2009. https://doi.org/10.1111/j.1468-2370.2008.00263.x.

3. Karatepe, O.M. “High-performance work practices and hotel employee performance: The mediation of work engagement.” International Journal of Hospitality Management, vol. 32, 2013, pp. 132–140. https://doi.org/ 10.1016/j.ijhm.2012.05.003.

4. Bakker, A.B. et al. “Using the job demands-resources model to predict burnout and performance.” Journal of Organizational Behavior, vol. 25, no. 3, 2004, pp. 293–315.

5. Mobley, W.H. “Intermediate linkages in the relationship between job satisfaction and employee turnover.” Journal of Applied Psychology, vol. 62, no. 2, 1977, pp. 237–240. https://doi.org/10.1037/0021-9010.62.2.237.

6. Tett, R.P. and J.P. Meyer. “Job satisfaction, organizational commitment, turnover intention and turnover: Path analyses based on meta-analytic findings.” Personnel Psychology, vol. 46, no. 2, 1993, pp. 259–293. https://doi. org/10.1111/j.1744-6570.1993.tb00874.x.

7. Schaufeli, W.B. et al. “The measurement of engagement and burnout: A two sample confirmatory factor analytic approach.” Journal of Happiness Studies, vol. 3, no. 1, 2002, pp. 71–92. https://doi.org/10.1023/A:1015630930326.

8. Allen, N.J. and J.P. Meyer. “Affective, continuance and normative commitment to the organization: An examination of construct validity.” Journal of Vocational Behavior, vol. 49, no. 3, 1996, pp. 252–276. https://doi. org/10.1006/jvbe.1996.0043.