A Novel Measure of "Good" Mentoring: Testing Its Reliability and Validity in Four Academic Health Centers.

Introduction: Despite the well-recognized benefits of mentoring in academic medicine, there is a lack of clarity regarding what constitutes effective mentoring. We developed a tool to assess mentoring activities experienced by faculty and evaluated evidence for its validity.

Methods: The National Initiative on Gender, Culture, and Leadership in Medicine-"C-Change"-previously developed the C-Change Faculty Survey to assess the culture of academic medicine. After intensive review, we added six items representing six components of mentoring to the survey-receiving help with career and personal goals, learning skills, sponsorship, and resources. We tested the items in four academic health centers during 2013 to 2014. We estimated reliability of the new items and tested the correlation of the new items with a mentoring composite variable representing faculty mentoring experiences as positive, neutral, or inadequate and with other C-Change dimensions of culture.

Results: Among the 1520 responding faculty (response rate 61-63%), there was a positive association between each of the six mentoring activities and satisfaction with both the amount and quality of mentoring received. There was no difference by sex. Cronbach α coefficients ranged from 0.89 to 0.95 across subgroups of faculty (by sex, race, and principal roles). The mentoring responses were associated most closely with dimensions of Institutional Support (r = 0.58, P < .001), Institutional Change Efforts for Faculty Support (r = 0.52, P < .001), Values Alignment (r = 0.58, P < .001), Self-efficacy (r = 0.43; P < .001), and Relationships/Inclusion/Trust (r = 0.41; P < .001).

Discussion: Data demonstrated that the Mentoring scale is a valid instrument to assess mentoring. Survey results could facilitate mentoring program development and evaluation.