The effects of negative mentoring experiences on protégés’ turnover intention: the roles of harmonious work passion and moqi with the mentor

Abstract

Purpose

This study tested a holistic model that investigated the interaction effect of negative mentoring experiences and moqi (pronounced “mò-chee”) with a mentor—where moqi refers to a situated state between two parties in which one party understands and cooperates well with the other party without saying a word—on the protégés’ turnover intention, along with the mediating role of protégés’ harmonious work passion.

Design/methodology/approach

Data were collected from 281 protégés through a three-wave questionnaire survey with a 1-month lag between waves. We used a hierarchical multiple regression and bootstrapping analysis to test our hypotheses.

Findings

Our results support the mediating effect of harmonious work passion on the positive relationship between protégés’ negative mentoring experiences and turnover intention. In addition, our analysis confirmed that moqi with the mentor amplifies both the impact of protégés’ negative mentoring experiences on harmonious work passion and the indirect effect of negative mentoring experiences on protégés’ turnover intention via harmonious work passion.

Originality/value

By demonstrating the interaction effect of protégés’ negative mentoring experiences and moqi with their mentor on turnover intention, as well as the mediating role of harmonious work passion, this study expands our understanding of the mechanism and boundary condition of the effect of negative mentoring experiences and provides inspiration and guidance for mentoring practices.

Citation

Hu, Z., Kwan, H.K., Zhang, Y. and Li, J. (2024), "The effects of negative mentoring experiences on protégés’ turnover intention: the roles of harmonious work passion and moqi with the mentor", Journal of Managerial Psychology, Vol. 39 No. 6, pp. 716-731. https://doi.org/10.1108/JMP-01-2023-0017

Introduction

Building a mentoring relationship between an experienced senior employee (mentor) and a less experienced junior employee (protégé) is commonly perceived as an effective means of employee development and retention (e.g. Zheng et al., 2021). However, mentoring relationships are not always as efficient as has been claimed (Eby et al., 2004, 2008; Topa and Perez-Larrazabal, 2016). Researchers have identified negative experiences in mentoring relationships that cause mentors and protégés to feel discordant, disappointed or dissatisfied (Eby et al., 2004). Negative mentoring experiences typically involve mentor–protégé mismatch, mentors’ inappropriate interactions with protégés (e.g. distancing behavior, manipulative behavior) and mentors’ dysfunctionality (e.g. lack of expertise, personal problems), and such experiences have been shown to have undesirable effects on protégés, leading to stress, depressed mood, psychological withdrawal at work, poor job satisfaction and turnover intention (Eby et al., 2004, 2010; Hu et al., 2022).

Despite their destructive impacts, negative mentoring experiences have not received sufficient attention from researchers or managers, particularly regarding their mediating mechanism and boundary conditions. One possible reason is that the frequency of negative mentoring experiences is low (Simon and Eby, 2003), with negative experiences simply regarded as minor aberrances in a mentoring relationship. As Eby et al. (2000), who first proposed the construct of negative mentoring experiences, stated, “we are not suggesting that … the presence of negative events means that the relationship is doomed to fail. Even in healthy relationships negative events occur …” (p. 2). However, it remains unclear whether negative mentoring experiences can be safely ignored when protégés get along well with their mentors. To clarify this issue, this study conceptualizes moqi (pronounced “mò-chee”) with the mentor—where moqi refers to a situated state between two parties in which one party understands and cooperates well with the other party without saying a word (Chen and Cole, 2022; Zheng et al., 2019)—as a boundary condition and investigates under which conditions (high versus low levels of moqi) negative mentoring experiences are most harmful.

In a Chinese setting, moqi consists of two characters: the character “mo,” which means “nonverbal and tacit,” and the character “qi,” which means “agreement and contracts for cooperation” (Zheng et al., 2019). In this study, we use moqi with the mentor, rather than mentor–protégé exchange or other variables related to the quality of mentoring relationships, to define the situation in which a protégé gets along with their mentor because moqi captures protégés’ cooperative behavior with their mentor in a tacit and contractual way (Zheng et al., 2019). We suggest that moqi with the mentor reflects deeper and more intrinsic aspects of mentor–protégé relationships than other variables related to the quality of such relationships (Zhang et al., 2022). Thus, the first purpose of this study is to empirically validate the potential interaction effect of protégés’ negative mentoring experiences and moqi with their mentor by investigating an important work outcome: turnover intention. Contrary to the ideal that negative mentoring experiences can be ignored in healthy mentor–protégé relationships, we suggest that negative mentoring experiences have a stronger positive effect on the turnover intention of protégés who have a higher level of moqi with their mentor.

In addition to examining the pivotal role played by moqi in mentoring relationships, we explore the more stable and deeper-level motivational mechanism that underlies the interaction effect mentioned above. We draw on self-determination theory (Deci and Ryan, 2000) and identify harmonious work passion as an effective motivational mechanism that mediates the joint effect of protégés’ negative mentoring experiences and moqi with their mentors on protégés’ turnover intention. Vallerand et al. (2003) indicated that work passion makes employees pleased and satisfied by enhancing their well-being and giving meaning to their work. Among passionate employees, harmoniously passionate employees are likely to be delighted and satisfied because harmonious passion results from the autonomous internalization of an activity rather than from intra- or interpersonal pressures (Houlfort et al., 2013). Although we acknowledge the potential relevance of intrinsic motivation in the context of negative mentoring experiences, we focus on harmonious work passion instead of other motivational conduits, such as intrinsic motivation at work, because harmonious work passion involves both an identification with and a feeling of love for work (Carpentier et al., 2012). Empirical studies have confirmed the view that the influence of harmonious passion on individuals’ turnover intention is above and beyond that of intrinsic motivation (Houlfort et al., 2013).

In summary, to better understand the mediating mechanism and boundary conditions of the effect of negative mentoring experiences, we propose a moderated mediation model to examine the mediating role of harmonious work passion and the moderating role of moqi with the mentor in the relationship between protégés’ negative mentoring experiences and protégés’ turnover intention. This study contributes to the literature in three ways. First, we address the relative dearth of research on the mechanism through which negative mentoring experiences influence protégés’ outcomes by investigating the mediating role of harmonious work passion. Second, this research contributes to our understanding of whether the detrimental effect of negative mentoring experiences varies depending on the quality of the mentor–protégé relationship by considering the moderating role of moqi with the mentor. Finally, from a practical standpoint, this study informs researchers and managers that negative mentoring experiences deserve more attention, given that when such experiences occur in relationships that are considered harmonious, they are even more harmful to protégés and their organizations than when they occur in less harmonious relationships.

Theory and hypotheses

Negative mentoring experiences and turnover intention

One of the main functions of a mentoring program is to train and retain talent in an organization. Research has provided abundant evidence of the effectiveness of mentoring relationships in reducing protégés’ turnover intention (see Eby et al., 2013, for a review). Unfortunately, mentoring programs do not always run smoothly and occasionally lead to negative mentoring experiences. These experiences are defined as events that bring about disharmony, dissatisfaction and disappointment between mentors and protégés and often manifest as a lack of affinity and similarity between mentors and protégés, as protégés’ perception of manipulative or distancing behavior by mentors or as mentors’ professional or personal issues (Eby et al., 2004).

Self-determination theory suggests that individuals have basic psychological needs for competence, relatedness and autonomy. When social conditions enable individuals to bond with another person while reinforcing their need for autonomy and competence, these individuals are motivated to stay and grow in such an environment (Broeck et al., 2010). In contrast, employees who cannot see the benefits of the environment or whose basic needs cannot be satisfied by the current conditions may want to leave. Given this, we suggest that a high-quality mentoring relationship provides protégés with information, opportunities and other scarce resources that are beneficial to their performance and career development. As a result, protégés are satisfied with their work status and willing to stay in their current work environment. In contrast, incompatibilities in mentor–protégé dyads, protégés’ perceptions of tyrannical or distancing behavior by mentors or professionally or interpersonally inept mentors are likely to impair closeness between protégés and mentors, impede protégés’ goal attainment and thus lead to increased turnover intention among protégés. Therefore, we propose the following hypothesis:

H1.

Protégés’ negative mentoring experiences are positively related to their turnover intention.

Mediating role of harmonious work passion

Work passion represents employees’ enthusiasm and excitement for their work. Passionate employees love their work, are willing to invest time and energy in it, and are likely to internalize work as part of their identity. On the basis of self-determination theory, Vallerand et al. (2003) proposed a dualistic model of work passion, dividing it into obsessive work passion and harmonious work passion. These two types of work passion stem from the different processes by which individuals internalize their work as part of their identity. Harmonious work passion is the result of employees’ autonomous internalization of work; that is, they recognize the importance of their work to them independently and internally, without being affected by external factors. In contrast, obsessive work passion stems from employees’ controlled internalization, in which they are forced by internal or interpersonal pressures (e.g. to gain self-esteem and recognition from others) to treat their work as part of themselves.

Managers and scholars have striven to improve employees’ work passion, especially their harmonious work passion, given that it has a positive and significant effect on a number of employee outcomes, such as greater positive affect, higher well-being, lower depression and less turnover intention compared with employees who lack harmonious work passion (Forest et al., 2012; Houlfort et al., 2013). Self-determination theory suggests that harmonious work passion is likely to be facilitated in a social context in which employees’ basic psychological needs for competence, relatedness and autonomy can be met (Deci and Ryan, 2000) because a needs-supportive context helps employees to internalize work autonomously into their self-concept. Furthermore, relationship motivation theory, a sub-theory of self-determination theory, examines the critical role of relationships expressly and posits that high-quality relationships can connect individuals to other people while satisfying their needs for autonomy and competence, and vice versa. In this sense, we expect negative mentoring experiences to impede protégés’ autonomous internalization of work and thus decrease their harmonious work passion.

Researchers have identified five types of negative mentoring experiences, namely mentor–protégé mismatch, mentors’ distancing behavior, mentors’ manipulative behavior, mentors’ insufficient expertise and general dysfunctionality. These experiences often hinder protégés’ access to task-related resources, recognition, opportunities for skill development and autonomy (Eby, 2007). First, dysfunctional mentors may establish barriers to their protégés’ success through distancing (e.g. consciously neglecting protégés, refusing to take protégés to important meetings) and manipulative behaviors (e.g. asking protégés to do chores, taking credit for their work) (Eby et al., 2004). Mentors may also fail to provide the necessary support and assistance for their protégés due to their lack of expertise, impeding protégés’ prospects and undermining their opportunities to achieve their goals and succeed at work (Broeck et al., 2010). Second, negative mentoring experiences, such as mismatch and disharmony between mentors and protégés, mentors’ poor attitudes and deviant behaviors toward protégés or mentors’ personal problems can undermine protégés’ trust and their sense of attachment to their mentors (Haggard, 2012) and their organizations (Eby et al., 2004). Finally, dysfunctional mentors are likely to exclude protégés from important tasks, assign them excessive or insufficient workloads and sabotage or take credit for protégés’ work. These experiences can undermine protégés’ ability to make choices, restrict their autonomy at work and generate psychological constraints. Given the detrimental effects mentioned above, protégés who have negative mentoring experiences perceive more work constraints and take less pleasure from their work than other protégés. They have difficulty internalizing work into their identity autonomously, which reduces their harmonious passion for work.

As mentioned, harmonious work passion stems from employees’ autonomous internalization of work, allowing them to devote themselves to work freely and voluntarily because of the work itself. As such, employees are free to determine and adjust their level of work involvement to ensure that their work will not clash with other demands outside of work. Thus, harmoniously passionate employees are likely to experience positive emotions at work (Carpentier et al., 2012) and, accordingly, they are willing to stay in their current job. Conversely, when employees are forced to immerse themselves in their work, they are likely to lose control of their work schedule and become mentally exhausted, which increases their intention to leave their current job. Studies have shown that harmonious work passion is strongly and negatively related to employee turnover intention (Gong et al., 2020; Houlfort et al., 2013); that is, harmoniously passionate employees are less likely than their counterparts to leave work because their basic needs are being met. Thus, we propose the following hypothesis:

H2.

Harmonious work passion mediates the positive relationship between protégés’ negative mentoring experiences and turnover intention.

Moderating role of moqi with the mentor

Moqi, which refers to a state in which one party can effectively perceive and interpret nonverbal and implicit signals from another party, is an important characteristic that can be used to describe the quality of a relationship, especially in an Eastern context (Zhang et al., 2022). Although moqi, in common with leader–member exchange (LMX) and followership (Sims and Weinberg, 2023), is a relational construct involving two parties, it differs from these variables. Specifically, Moqi is characterized by an understanding from only one party without excessive verbal communication, whereas LMX emphasizes exchanges between the parties in leader–follower relationships without specifically focusing on the direction of understanding. In addition, LMX and followership primarily pertain to leader–follower relationships, whereas moqi can be observed in various interpersonal relationships, such as peer relationships, friendships and mentor–protégé relationships (as examined in our study), and can serve as a valuable resource for fostering high-quality relationships (Li et al., 2020). We chose to focus on moqi rather than LMX because our study focuses on mentor–protégé relationships. Furthermore, our study examines the perspective of protégés alone, considering whether they perceive an intimate connection with their mentor in their own psychological realm, without necessarily considering the existence of reciprocal benefits between the two parties.

Moqi with the mentor represents the extent to which protégés interact and cooperate well and tacitly with their mentors (Zheng et al., 2019). Empirical evidence has indicated that moqi can improve information transfer efficiency, strengthen communication effectiveness, enhance employees’ understanding of their task objectives and boost subordinates’ trust in their supervisor (Chen and Cole, 2022; Li et al., 2020; Zheng et al., 2019). Ideally, negative mentoring experiences would not be detrimental when protégés have a high level of moqi with their mentor. However, in this study, we argue that the opposite, proposing that moqi with the mentor increases the detrimental impact of negative mentoring experiences on protégés’ harmonious work passion for two reasons.

First, according to self-determination theory, individuals inherently have basic needs for competence, relatedness and autonomy, and they tend to seek support to satisfy these needs from their social context, such as through social relationships (Broeck et al., 2010). A high level of moqi with their mentors can help protégés develop positive feelings and build a deep interpersonal bond with their mentors (Chen and Cole, 2022), which will increase their dependence on their mentors. The more that protégés rely on their mentors for social support, the more likely they are to be affected by their mentors (Wee et al., 2017). From this perspective, compared with protégés with a low level of moqi with their mentors, protégés who have a high level of moqi with their mentors are more likely to trust their mentors and rely on them to satisfy their basic needs; as a result, these protégés will be more affected by negative mentoring experiences than their counterparts.

Second, the presence of ambivalence can aggravate the negative effect of negative mentoring experiences. Ambivalence refers to simultaneous and contradictory feelings toward an object or person (Ashforth et al., 2014). The interaction of positive and negative experiences (i.e. moqi with the mentor and negative mentoring experiences in this study) has a stronger negative effect on individuals than negative experiences alone, as ambivalence is often aversive (Methot et al., 2017). A high level of moqi with the mentor enhances protégés’ trust in their mentor and generates positive expectations regarding the latter’s intentions and behaviors (Li et al., 2020). In this case, when negative mentoring events occur, protégés tend to feel that their trust in and expectations of their mentor have been violated, leading them to feel unbalanced, confused, anxious and overwhelmed (Methot et al., 2017). These negative emotions and perceptions make it difficult for protégés to meet their basic psychological needs. In addition, the interaction of moqi with negative experiences creates unpredictability, which can make protégés feel out of control at work and reduce their autonomy (Pratt and Doucet, 2000), thus increasing the negative effect of negative mentoring experiences on their harmonious work passion. Therefore, we propose the following hypothesis:

H3.

Protégés’ perceived moqi with their mentors moderates the relationship between their negative mentoring experiences and their harmonious work passion, such that this relationship is stronger when the level of moqi is high rather than low.

Integrating Hypothesis 2 (which predicts the mediating role of harmonious work passion in the relationship between protégés’ negative mentoring experiences and their turnover intention) and Hypothesis 3 (which predicts the moderating effect of moqi with the mentor in the relationship between protégés’ negative mentoring experiences and harmonious work passion), we develop a moderated mediation model in which moqi with the mentor moderates the indirect effect of protégés’ negative mentoring experiences on their turnover intention via harmonious work passion. Thus, we propose the following hypothesis:

H4.

Protégés’ perceived moqi with their mentors moderates the indirect relationship between negative mentoring experiences and turnover intention via harmonious work passion, such that the indirect effect is stronger when their level of moqi is high rather than low.

Method

Sample and procedures

To test our hypotheses, we designed a nonexperimental study with surveys. Surveys are well-suited to capture the intricate relationships and can provide valuable insights into the interplay between negative mentoring experiences and employee turnover intention in a realistic organizational context. However, the nonexperimental design has inherent limitations in terms of establishing causal relationships (Shadish et al., 2002). To enhance the reliability and validity of the findings, we measured the predictor, mediator and outcome variables at three different time points (Law et al., 2016).

The data were collected from 281 protégés employed in three state-owned enterprises. To obtain reliable protégé–mentor dyad information, we first contacted the human resources managers at each enterprise. Once the protégé–mentor dyads were identified, research assistants visited the enterprises in person, where they formally communicated the purpose of the study in different sessions and invited the protégés to participate in the study while ensuring the confidentiality of their responses. The protégés who volunteered to participate were asked to complete surveys using the paper-and-pencil method during their working hours at three time points (Time 1, Time 2 and Time 3), each separated by 1 month.

At Time 1, the protégés were asked to provide information about their demographic characteristics (i.e. age, gender, tenure, education level and position), negative mentoring experiences when interacting with their mentor and level of moqi with their mentor. Initially, 465 protégés agreed to participate in the study, and we received 442 complete questionnaires at Time 1. One month later (Time 2), the research assistants contacted only the protégés who had completed the survey at Time 1 and invited them to rate their harmonious work passion. At Time 2, of the 442 protégés, we received 313 complete questionnaires. One month later (Time 3), the 313 protégés who completed the survey in the previous two phases were asked to rate their intention to leave their company. Next, we used each protégé’s hidden code to match their responses collected at the three time points. We obtained 281 matched and valid questionnaires from the 465 protégés, for an overall response rate of 60.43%. Monte Carlo simulations were employed to verify the adequacy of our sample size for testing the proposed moderated mediation model (Schoemann et al., 2017). The results revealed that with a sample size equal to or greater than 280, the statistical power reached 0.99, substantiating that the sample size in this study (n = 281) is adequate for testing our theoretical model.

Measures

Protégés’ negative mentoring experiences (Time 1). The five-dimension scale developed by Eby et al. (2004) was used to measure the protégés’ negative mentoring experiences. The five dimensions are (1) mismatch within the dyad (nine items, including “My mentor and I have dissimilar personalities”); (2) distancing behavior (seven items, including “My mentor keeps me ‘out of the loop’ on important issues”); (3) manipulative behavior (11 items, including “My mentor has deliberately misled me”); (4) lack of mentor expertise (seven items, including “I have my doubts about my mentor’s job-related skills”) and (5) general dysfunctionality (eight items, including “My mentor tends to bring his/her personal problems to work”) (α = 0.98). A 7-point Likert-type scale was used, ranging from 1 (never) to 7 (always).

Moqi with the mentor (Time 1). Moqi with the mentor was measured using an eight-item scale adapted from Zheng et al. (2019). The original scale was developed to measure subordinates’ moqi with their supervisor, and we used this scale to measure the protégés’ moqi with their mentor after appropriate modifications. We asked the protégés to assess their perceived level of moqi with their mentor, considering the following instruction, “In day-to-day work situations, without explicit verbal communication or overt cues from my mentor,” and the following sample items, “I cooperate well with my mentor” and “I am able to understand his/her train of thought” (α = 0.89). A 7-point Likert-type scale was used, ranging from 1 (strongly disagree) to 7 (strongly agree).

Harmonious work passion (Time 2). To measure harmonious work passion, we adopted the seven-item harmonious passion scale developed by Vallerand et al. (2003). Sample items are “My work allows me to live memorable experiences” and “My work is in harmony with the other activities in my life” (α = 0.96). A 7-point Likert-type scale was used, ranging from 1 (strongly disagree) to 7 (strongly agree).

Turnover intention (Time 3). Turnover intention was measured with three items from Konovsky and Cropanzano (1991). A sample item is “I often think about quitting my job at [company name]” (α = 0.83). A 7-point Likert-type scale was used, ranging from 1 (strongly disagree) to 7 (strongly agree).

Control variables. According to the literature (Eby et al., 2004; Wright and Bonett, 2007), surface-level demographic variables and characteristics of mentoring relationships may provide alternative explanations for or influence the relationships proposed in this study. Therefore, we controlled for the participants’ demographic variables (i.e. protégé age, protégé gender, protégé education level, protégé tenure and protégé position) and diversity in the mentoring relationship (i.e. mentor–protégé gender difference and age difference). Because the participants in this study were recruited from three companies, we also controlled for two company dummy variables to exclude the possible influence of company differences.

Results

Confirmatory factor analysis

We conducted a set of single-level confirmatory factor analyses (CFAs) using Mplus 8.4 to test the discriminant validity of the four focal latent variables in our research model. To account for possible standard errors due to the large number of items measuring negative mentoring experiences and the number of parameters to be estimated, the 42 items of negative mentoring experiences were packaged into five second-level items (Little et al., 2002). The results showed that our hypothesized four-factor model fitted the data well: χ² = 804.969, df = 224, p < 0.001, CFI = 0.910, TLI = 0.899, RMSEA = 0.096 and SRMR = 0.065. In addition, as the four-factor model performed significantly better than the other alternative models, the CFA results indicated that the four variables had good discriminant validity.

Descriptive statistics

Table 1 reports the means, standard deviations and correlations of the study variables.

Hypothesis testing

Hayes’s (2013) PROCESS macro for SPSS was used to test our hypotheses. The results are summarized in Table 2. Models 1 and 4 tested the effects of the control variables on harmonious work passion and turnover intention, respectively. Hypothesis 1 posits a positive association between protégés’ negative mentoring experiences and their turnover intention. The results of Model 5 reveal that negative mentoring experiences were positively related to turnover intention (β = 0.683, p < 0.001). Thus, Hypothesis 1 is supported.

Hypothesis 2 postulates the mediating effect of harmonious work passion on the positive relationship between protégés’ negative mentoring experiences and turnover intention. Our bootstrapping results (20,000 resamples) showed that the indirect effect of negative mentoring experiences on turnover intention via harmonious work passion was significant (b = 0.103, 95% confidence interval [CI] = [0.046, 0.196]). Thus, Hypothesis 2 is supported.

Hypothesis 3 predicts the moderating effect of protégés’ perceived moqi with their mentor on the relationship between negative mentoring experiences and harmonious work passion. The results of Model 3 in Table 2 show that the interaction term of protégés’ negative mentoring experiences × moqi with the mentor was statistically significant (β = −0.125, p < 0.01), indicating that the negative effect of negative mentoring experiences on harmonious work passion was amplified by the level of moqi with the mentor, providing support for Hypothesis 3. Furthermore, we used a simple slope analysis to assess the conditional effects of high and low levels of moqi with the mentor (1 standard deviation above and below the mean) on harmonious work passion (see Figure 1). The results indicate that the negative relationship between negative mentoring experiences and harmonious work passion was more pronounced at a high level of moqi with the mentor (β = −0.151, p < 0.01) than at a low level (β = 0.028, n.s.), further supporting Hypothesis 3.

Finally, Hypothesis 4 predicts that the indirect effect of negative mentoring experiences on turnover intention through harmonious work passion varies at different levels of moqi with the mentor (1 standard deviation above and below the mean). The results of bootstrapping analyses (20,000 resamples) show that when the level of moqi with the mentor was high, the indirect effect of negative mentoring experiences on turnover intention via harmonious work passion was significant (indirect effect = 0.099, 95% CI = [0.027, 0.213], excluding 0), whereas it was not significant when the level of moqi with the mentor was low (indirect effect = −0.018, 95% CI = [−0.063, 0.068], including 0). Accordingly, these results confirm the presence of moderated mediation, supporting Hypothesis 4.

Discussion

Studies have confirmed that mentorship is not always effective and that negative mentoring experiences are not rare for either mentors or protégés (Eby et al., 2004, 2008). Researchers and managers have verified the negative impact of negative mentoring experiences to some extent, but we still know little about the mediating mechanism and boundary conditions. Thus, this study examined the effect of negative mentoring experiences on protégés’ turnover intention, along with the mediating effect of harmonious work passion and the moderating effect of moqi with the mentor. Our empirical results show that negative mentoring experiences blunt protégés’ harmonious work passion, increasing their intention to leave their company. This effect was stronger for protégés with a high (versus low) level of moqi with their mentor.

Theoretical implications

First, our results confirm the significant association between negative mentoring experiences and turnover intention by revealing the key role of mentorship in satisfying protégés’ psychological needs. As one of the most important types of interpersonal relationships for protégés in a work context, mentoring relationships can satisfy protégés’ basic psychological needs by providing them with challenging tasks, opportunities to showcase their talent, training and protection (i.e. career development support), as well as unconditional recognition and affirmation, kindness and advice (i.e. social psychological support) (Kram, 1985). Accordingly, when dysfunctional mentoring events occur, protégés find that their basic needs are not satisfied, which gradually leads to a desire to leave their current company. Our study verifies the detrimental influence of negative mentoring experiences on protégés’ work-related outcomes and offers a new perspective for future studies of negative mentoring experiences.

Second, this study identifies the mediating role of harmonious work passion in the relationship between negative mentoring experiences and turnover intention, thus opening the “black box” of the detrimental effect of negative mentoring experiences on protégés’ outcomes. Self-determination theory points out that work passion, especially harmonious work passion, is a fundamental psychological bridge between the social environment (including leaders, colleagues, relatives and friends) and individual work-related outcomes (Forest et al., 2012). The social environment can promote or undermine individuals’ internalization of external motivation by satisfying or failing to satisfy their needs for competence, relatedness and autonomy (Deci and Ryan, 2000). Our findings reveal that harmonious work passion is also a critical motivational process explaining the impact of negative mentoring experiences on protégés’ turnover intention, as mentoring relationships are among the closest and most influential relationships in the workplace. Therefore, we encourage researchers to further explore the important role of harmonious work passion in mentoring relationships to better understand the deep mechanism of the influence of negative mentoring experiences.

Third, Eby et al. (2004) suggested that negative mentoring experiences are not equivalent to negative mentoring relationships, as the former are only episodic and rare events in a mentoring relationship. Therefore, some scholars and practitioners have questioned the importance of negative mentoring experiences and the need to pay attention to them (Eby, 2007). Our study introduces an ambivalent perspective on exploring the moderating effect of moqi with the mentor on the effect of negative mentoring experiences. Our results show that for protégés with a high level of moqi with their mentor, negative mentoring experiences may have a stronger negative impact than for those with a low level of moqi with their mentor. This finding refutes the stereotype that negative mentoring experiences can be ignored in a healthy mentoring relationship and has implications for future exploration of the negative aspects of mentoring relationships.

Practical implications

First, studies have shown that mentoring relationships should be used as an important management tool to train and retain valued employees, whereas the findings of the current research suggest that this tool may backfire in the event of negative mentoring experiences. Thus, this study reconfirms the harmfulness of abnormal events or mentors’ dysfunctional characteristics, and argues that managers should pay attention to them. In addition, potential protégés should be reminded that mentoring relationships are intended to be beneficial and fulfilling, although conflicts and difficulties may be inevitable. Managers should provide timely guidance and counseling to protégés who have had a dysfunctional mentoring relationship to prevent further negative consequences.

Second, in view of the key role of harmonious work passion in the relationship between negative mentoring experiences and turnover intention, managers should adjust their strategies for selecting and training mentors, and encourage and cultivate mentors to provide support for autonomy rather than support that results in controlling their protégés; thus, managers should provide proactive and open communication between mentors and protégés, a flexible work design and empowerment of their protégés. In this way, an independent work environment can be created for protégés to cultivate their harmonious work passion and thus increase their intention to stay in their current job.

Third, although negative mentoring experiences are low-frequency events in the actual practice of mentoring programs, managers must remain vigilant. Even if mentors and protégés usually get along and cooperate well, managers must continue to pay attention and provide timely support and guidance when problems are detected. Indeed, our results showed that for consistently harmonious mentoring relationships, the consequences of negative mentoring experiences are more serious than for less harmonious mentoring relationships.

Finally, considering the growing body of research on workplace coaching in recent years and the similarities between coaching and mentoring relationships (Jones et al., 2016), our study’s findings have implications for research in the coaching field. Although coaching relationships may not be as long-term as mentoring relationships (Brockbank and McGill, 2006), negative experiences that arise during interactions between coaches and “coachees” (i.e. those supported by the coach) are equally deserving of attention. Particularly for coaches and coachees who already share mutual trust and rapport, these negative experiences may appear as sporadic incidents but can significantly impact the ultimate achievement of coaching goals.

Limitations and suggestions for future research

This study has two limitations. First, as the data on negative mentoring experiences were reported by the protégés themselves, it is unclear whether their perception of dysfunctional mentoring reflected a real dysfunctional mentoring relationship or simply a subjective interpretation of their discontent with their organization or job. We encourage researchers to collect data from both protégés and mentors in future studies to further examine the effect of negative mentoring experiences and provide an objective and comprehensive understanding of negative and dysfunctional mentoring relationships. Second, although our results reveal that negative mentoring experiences can affect harmonious work passion and increase turnover intention, our time-lagged research design may have failed to provide sufficient evidence to support the causal relationships between these variables. Indeed, we cannot rule out the possibility that the protégés’ desire to quit their job (for other reasons) reduced their passion for work, leading them to report more negative mentoring experiences. According to social exchange theory, individual behavior at work is driven by “exchange” relationships. Protégés may consider quitting because they have not received their expected rewards (such as salary, promotion opportunities and social support) to the extent that they no longer expect to receive them. As a result, they lose motivation to continue investing in their work, leading to a decrease in work passion.

In addition, there are other unmeasured mediators, such as ego depletion (Hu et al., 2022) and organizational commitment (Payne and Huffman, 2005), that may also play a mediating role in this process. This study serves as a springboard, and future research should test our hypotheses using vignette experiments or a longitudinal research design to further verify the relationships proposed in our study.

Figures

Figure 1

The interactive plot

Table 1

Means, standard deviations and correlations

Variables	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12
1. Protégé gender	1.250	0.433
2. Protégé age	27.380	5.959	−0.002
3. Protégé education level	3.620	0.993	−0.019	−0.335**
4. Protégé job tenure	5.069	6.238	−0.015	0.891**	−0.532**
5. Protégé position	1.110	0.360	−0.068	0.110	−0.108	0.170**
6. Firm1	0.370	0.484	−0.050	−0.394**	0.488**	−0.492**	−0.181**
7. Firm2	0.210	0.408	0.148*	0.308**	−0.463**	0.356**	0.347**	−0.395**
8. Mentor–protégé gender difference	0.006	0.490	0.581	−0.071	−0.010	−0.049	−0.125	0.036	−0.011
9. Mentor–protégé age difference	8.771	7.100	−0.013	−0.340	0.151	−0.297	−0.111	0.256	−0.088	0.035
10. Protégé negative mentoring experiences	2.261	1.094	−0.046	0.001	−0.025	0.072	0.099	0.028	0.103	−0.036	−0.061
11. Harmonious work passion	4.128	0.768	−0.089	0.054	−0.011	0.062	0.084	−0.138*	−0.078	0.024	−0.130*	−0.245**
12. Turnover intention	2.638	1.437	−0.036	−0.104	−0.048	−0.062	0.068	0.132*	0.181**	−0.073	0.126*	0.534**	−0.453**
13. Moqi with the mentor	4.256	0.720	−0.090	0.134*	0.072	0.083	0.107	−0.012	−0.117	0.004	−0.076	−0.333**	0.606**	−0.327**

Note(s): N = 281. Protégé gender was coded as 0 = male, 1 = female. Protégé age and job tenure were coded in years. Protégé education level was coded as 1 = junior high school or below, 2 = high school, 3 = associate degree, 4 = bachelor’s degree and 5 = master’s degree or above. Protégé position was coded as 1 = first-line employees, 2 = first-line managers, 3 = middle managers, 4 = top managers and 5 = CEO

*p < 0.05, **p < 0.01. Two-tailed tests

Source(s): Authors’ own creation

Table 2

Results of hierarchical regression analysis

Variables	Harmonious work passion			Turnover intention
	Model 1	Model 2	Model 3	Model 4	Model 5
Control variables
Protégé gender	−0.252(0.131)	−0.273*(0.127)	−0.124(0.104)	−0.057(0.240)	0.023(0.203)
Protégé age	0.002(0.019)	−0.012(0.018)	−0.023(0.015)	−0.056(0.034)	−0.002(0.029)
Protégé education	0.020(0.064)	0.044(0.062)	0.018(0.051)	−0.057(0.117)	−0.145(0.099)
Protégé job tenure	0.000(0.002)	0.001(0.002)	0.001(0.001)	0.003(0.003)	−0.002(0.003)
Protégé position	0.225(0.137)	0.242(0.133)	0.051(0.111)	0.023(0.251)	−0.041(0.213)
Firm1	−0.290*(0.116)	−0.235*(0.113)	−0.202*(0.093)	0.683**(0.212)	0.476**(0.181)
Firm2	−0.303*(0.139)	−0.228(0.136)	−0.080(0.112)	0.971***(0.256)	0.688**(0.218)
Mentor–protégé gender difference	0.201(0.114)	0.197(0.111)	0.074(0.091)	−0.232(0.210)	−0.218(0.177)
Mentor–protégé age difference	−0.010(0.007)	−0.013(0.007)	−0.009(0.005)	0.014(0.013)	0.024*(0.011)
Independent variable
Protégé negative mentoring experiences		−0.180***(0.041)	−0.062(0.036)		0.683***(0.066)
Moderator
Moqi with mentor			0.625***(0.055)
Interaction
Protégé negative mentoring experiences × Moqi with the mentor			−0.125**(0.039)
Mediator
Harmonious work passion
R2	0.074	0.136	0.430	0.112	0.366
ΔR2		0.061***	0.294***		0.253***
F	2.415*	4.239***	16.826***	3.805***	15.566***

Note(s): N = 281. *p < 0.05, **p < 0.01, ***p < 0.001. Two-tailed tests

Source(s): Authors’ own creation