آیا تجارب بد قوی تر از تجارب خوب در روابط مشاوره هستند؟ مدارک و شواهد ازدیدگاه شاگرد و مشاور

Two studies examined the relative importance of good versus bad mentoring experiences in predicting subjective states associated with the mentoring relationship. Study 1 examined the protégé perspective and found general support for the proposition that, on average, bad is stronger than good in predicting protégé outcomes. Study 2 adopted the mentor perspective and found mixed support for the prediction that, on average, bad is stronger than good. The results are discussed in terms of advancing research and theory on the relational processes associated with mentoring in the workplace and the need to consider the relational context to more fully understand the relative predictive power of good and bad mentoring experiences.

Four overall conclusions can be reached from Study 1 and Study 2. First, bad is generally stronger than good in terms of predicting protégé outcomes. Second, the evidence is mixed in terms of bad being stronger than good in predicting mentor outcomes. Third, a differential pattern of effects for bad versus good exists in predicting relationship quality for protégés and mentors. Fourth, the relative predictive power of good and bad for well-being varies for mentors and protégés. Taken together these studies illustrate the importance of examining both good and bad mentoring experiences, and doing so from both the protégé and the mentor perspective. Comparing the predictive power of bad versus good for protégés and mentors For both protégés and mentors bad experiences outweighed good ones in predicting intent to stay in the relationship. This is consistent with Baumeister et al. (2001) theorizing that bad experiences carry more weight in predicting relationship stability than do comparable good experiences. In fact, Gottman's research on romantic relationships finds that good interactions must outnumber bad ones by a ratio of at least five to one for a relationship to stay intact. If the ratio of good to bad experiences falls below this level, then the relationship is likely to break-up (Gottman, 1994). It may also be that simply having a weak bond with one's partner is not sufficient to initiate a break-up since there are costs associated with leaving (Huston & Burgess, 1979). These costs can be amplified in situations where alternative relational partners are limited, a power difference exists between individuals, or the relationship is highly public (Levinger, 1979), as are the case with many workplace mentoring relationships. Differences were found across the two samples in terms of both perceptions of relationship quality and well-being. For protégés, overall good outweighed overall bad in terms of predicting relationship quality, yet the opposite effect existed for mentors (i.e., bad > good). This may reflect the asymmetrical nature of mentoring relationships. Although the mentor may benefit from the relationship, the primary goal is protégé growth and development (Eby et al., 2007 and Kram, 1985). As such, protégés may reap additional higher-order positive benefits when in a high quality mentoring relationship, such as enhanced career self-efficacy and career success. This is less likely to be the case for mentors. The relative salience of bad mentoring experiences in predicting mentors' perceptions of relationship quality may reflect sensitivity to the investment of time and energy in a relationship if their efforts are not appreciated or attempts to foster protégé grow and develop seem futile. In other words, for mentors the quality of the relationship may be judged more in terms of whether their own efforts to help the protégé are thwarted by relational problems. For protégés, bad was significantly more predictive than good in predicting well-being, whereas no significant difference was found for mentors. One explanation is that protégés are particularly vulnerable in a mentoring relationship. As such, bad experiences may be more damaging to a protégé's self-esteem and create concerns regarding how bad experiences with a mentor will influence day-to-day work experiences or even hinder future career prospects (Scandura, 1998). This may induce psychological strain in a protégé in the form of depressed mood or psychological withdrawal. While mentors may be affected by the stress associated with bad experiences with a protégé, the potential threat is far less for mentors since protégés do not control valued resources for mentors and mentors can easily distance themselves from protégés by simply reducing their engagement in the relationship. Overall, our findings suggests that the relative balance of good and bad experiences in relationships needs to be assessed not in a vacuum, but in the context of the relationship itself. The impact of a relationship appears to depend on one's role, and mentoring scholars have long recognized that the costs and benefits associated with being a mentor differ from that of the protégé. This helps explain why the relative weight of good and bad relational experiences differ for protégés and mentors. Implications for practice There are several practical suggestions based on our findings. It seems important to inform potential mentors and protégés of the full range of relational experiences they might encounter in a mentoring relationship. Such educative efforts are important since the popular press tends to present mentoring as an essential ingredient for protégé development, yet we know that often mentors and protégés have unmet expectations in mentoring relationships (Eby and Lockwood, 2005, Young and Perrewé, 2000a and Young and Perrewé, 2000b). In addition, given the potency of bad mentoring experiences for protégés, mentoring may be best suited for organizational contexts where there is top management support for mentoring and perceived accountability among mentors. An organizational climate supportive of mentoring is related to both mentor reports of relational quality and protégé reports of receiving more mentoring (Eby, Lockwood & Butts, 2006). Moreover, protégés' perceptions of mentor accountability are related to fewer bad mentoring experiences as reported by protégés (Eby, Lockwood & Butts, 2006). Thus, organizational support for mentoring and formal sanctions to deal with inappropriate mentor behavior may mitigate the effects of negative mentoring experiences for protégés and help shift focus to maximizing benefits of positive mentoring experiences. In terms of formal mentoring programs, although previous research demonstrates the importance of mentor and protégé training (Allen et al., 2006a and Allen et al., 2006b), little is known about what to emphasize in training sessions. Our results suggest that topics such as the pros and cons of mentoring relationships, how to set realistic relationship expectations, trust-building, effective conflict management, and how to recognize problems in the mentorship may be important to include, especially for protégés. Our results also highlight the importance of having procedures in place to allow one or both individuals to exit a mentoring relationship without negative repercussions since bad mentoring experiences contributed significantly to the prediction of intentions to exit the relationship. This has been offered as a recommendation for the design of formal mentoring programs (Allen, Finkelstein & Poteet, 2009) and our findings confirm the importance of having such a system in place, particularly for protégés. Limitations and directions for future research Like all research the present research has several limitations. The cross-sectional nature of the present research precludes strong cause-and-effect conclusions. While reverse causation is a possibility for some of the predicted associations, both theory and logic argue that relational experiences are likely to influence subjective states associated with the relationship. Nonetheless, mentoring relationships are dynamic (Eby et al., 2007), and as a consequence, reciprocal effects may also exist. For example, good relational experiences may engender positive reactions to the mentorship, which in turn increases the probability that good relational experiences will occur in the future. We also did not examine mediating variables to determine how and why negative experiences may relate to various protégé or mentor outcomes. For example, negative experiences with mentors may erode protégé self-esteem or engender fear of retaliation, which in turn may create psychological strain (Eby and Allen, 2002 and Scandura, 1998). Different process-oriented variables may be operating for mentors. For instance, negative experiences with protégés may lead to mentor frustration or reduced commitment to the relationship, which in turn reduces relationship quality. A related limitation is our focus on individual effects, which does not take into consideration how events experienced by one partner may affect the other partner. Since relationships may become mutually reinforcing (or punishing) over time, it is possible that the occurrence of some mentoring experiences, bad and good, are due to quid pro quo behaviors precipitated by the other member. We know of no published mentoring research using a longitudinal, dyadic research design and realize the difficulty associated with collecting such data. Nonetheless, in order to fully understand mentor–protégé relational exchanges it will be important to move mentoring scholarship this direction. We also recognize that the decision to focus on average good and bad relational experiences does not provide a nuanced understanding of how specific relational events influence the relationship. There are two specific concerns here. One concern is that there may be a critical incident or a specific relationship tipping point whereby one highly salient bad (or good) relational experience irrevocably alters the course of the relationship. Examining average relational experiences masks any such effects. Another concern is that we did not examine the relative predictive power of specific good and bad mentoring experiences to see if certain types of experiences were particularly potent predictors of protégé or mentor outcomes. For example, Eby (2007) suggests that bad mentoring experiences exist on a continuum where some experiences (e.g., manipulative behavior by the mentor) are more personally and professionally damaging than are others (e.g., mentor–protégé mismatches). Likewise, it may be that different experiences are more or less predictive of outcomes, depending on the specific outcome examined. For instance, the good experience of feeling personal satisfied by helping a protégé grow and develop may be a stronger predictor of mentor perceptions of relationship quality than the receipt of help on the job from the protégé. We encourage researchers to build on our results by examining the relative predictive power of specific good and bad experiences. Our findings are also bound by the choice of dependent variables. Other mentoring outcomes such as work attitudes, career perceptions, and career success may be related to good and bad mentoring experiences in different ways. We chose to focus on subjective states associated with the relationship in an effort to extend existing research on mentoring and answer recent calls for researchers to expand the criterion space of mentoring outcomes. Additional research is needed which examines the relative contribution of good and bad mentoring in predicting other relevant mentor and protégé outcomes at the individual and dyadic level of analysis. We should also note that although statistically significant, good and bad mentoring did not account for much variance in indicators of psychological well-being, particularly for mentors. Therefore, conclusions reached regarding the relative predictive power of good and bad mentoring needs to be tempered by this finding. In fact, our failure to find significant differences in mentor burnout between good and bad mentoring may be a function of the small amount of variance accounted for (only 4%) along with elevated standard errors, both of which make it more difficult to obtain statistical significance. Finally, common method variance is a potential concern. The hypotheses examined in the present study required either the protégé or the mentor to be the informant, given the subjective and personal nature of the constructs under study. In addition, the use of RWA should make this less of a concern in the present study since this data analysis procedure is designed to parse out unique effects by creating orthogonal predictor variables through a statistical transformation process. In closing, the present research extends mentoring scholarship by examining the relative predictive power of good versus bad mentoring experiences for both protégé and mentor outcomes. Our findings illustrate the importance of considering the full range of mentoring experiences in order to provide a comprehensive perspective on mentoring, as well as the utility of examining relational processes from both the mentor and the protégé perspective. We hope that this research serves as a springboard for future research on the complex dynamics associated with mentoring relationships in the workplace.