The mentor-protégé affinity on mentoring outcomes: the mediating effect of developmental networking

Abstract

Despite increasing role of mentoring in public organizations, mentoring and its outcomes have been under-investigated in public management. The purpose of the study is to fill the void by focusing on the mentor-protégé relationship and its effect on mentoring outcomes. Given the complexity of psychological interactions between mentor and protégé, this study examined the influence of the mentor-protégé affinity on mentoring outcomes both directly and indirectly through developmental networking. The results found that the mentor and protégé affinity did not have a direct impact on mentoring outcomes. Instead, its effects were fully mediated by internal and external developmental networking, which were the direct predictors of mentoring outcomes. This implies that finding a good match between mentor and protégé is important, but it is effective to mentoring outcomes only through the establishment of developmental networking.

Keywords:

• mentor-protégé affinity
• developmental networking
• mentoring outcome
• public organizations

Notes

1. These variables were borrowed from Feeney and Bozeman (2008) and Bozeman and Feeney (2009b). This article recognizes that there are many other mentoring outcomes such as salary increase, career commitment and turnover intention. In particular, salary is the most frequently used indicator of objective mentoring outcomes because ‘it better reflects the economic value an organization assigns for employees and their performance’ (Wolff & Moser, 2009, p. 205). Unfortunately, salary and many other mentoring outcome data were not available from the NASP-III dataset. The data limitation is also one of the reasons why we used protégé-now-mentor as a surrogate measure of career advancement, although taking a role of mentor could be a more personal choice if it is not officially required in public organizations.

2. Bowlby (1979) described four phases of attachment development during early childhood: indiscriminate sociability, attachment in the making, clear-cut attachment and reciprocal relationships. First, indiscriminate sociability refers to an infant’s response to cries, smiles and gazes indiscriminately to engage contact and affection from anyone nearby. Second, attachment in the making refers to an infant’s behaviors that become associated with a specific caregiver. Third, clear-cut attachment begins when an infant shows anxiety with separation from caregivers and in the presence of unfamiliar people. Finally, child and caregiver develop mutual appreciation and influence in reciprocal relationships.

3. Similarity-attraction theory posits that people are attracted to others who are similar to themselves in terms of personality, physical characteristics, socioeconomic status, religious beliefs and even bad habits (Berscheid & Walster, 1969; D. Byrne 1971). In a similar vein, Ragins (1997) found that the higher the similarity between mentor and protégé, the higher the likelihood of the mentor providing both psychological and career support. This perception of similarity between mentor and protégé tends to develop nurturing relationships, in which both parties experience mutual understanding and respect as they continuously involve themselves in the learning cycle (Allinson, Armstrong & Hayes, 2001). In a large, formal teacher mentoring program in New York City, for example, interpersonal similarities between mentors and protégés increased the protégés’ career satisfaction and influenced the decision to continue their job as a public school teacher (Owen & Solomon, 2006). Nonetheless, other scholars argue that people may be more likely to be attracted to those who have dissimilarity rather than similar attributes (e.g., Rosenbaum, 1986). Dissimilarity or diversity might contribute to personal development to a greater extent than similarity or homogeneity.

4. Using promotion as a dependent variable may raise a concern about causal ordering, because the variable asks whether the current job was a promotion in the current organization. Although the causal ordering of the variable may be a problem, we still keep the variable for the following two reasons: (1) more than 88% (352 of 406 sample respondents) of respondents said that their current jobs are not their first jobs, and (2) almost 80% of respondents started mentoring before the year 2000, about six years before the data were collected. Given the time and causation issues of the promotion variable, we included ‘protégé-now-mentor’ as another key variable to measure career advancement to cross-validate the findings.

5. AMOS does not support a 99.9% credible interval to test our hypotheses at the 0.001 significance level. Table 4 only shows the results of Bayesian estimation using a 95% creditable interval.