gms | German Medical Science

Background: The success of any mentoring program builds on the quality and durability of its mentor-mentee-relationships. To achieve a high standard in mentoring relationships, the process of matching mentors to their mentees plays a key role. In developing a mentoring program for medical students we addressed the need for an innovative matching method at a faculty with a large number of mentors and mentees. In order to automate and optimize the matching process as much as possible we developed a match-making algorithm. In the evaluation of the project we assessed whether the matching procedure yields mutually satisfying mentoring relationships.

Methods: Before being able to match, each mentor and mentee is required to create an online matching profile, covering career as well as professional, social and cultural preferences. The profile consists of questions with a 6-point Likert scale, choices of medical specialty and additional qualifications, and a mandatory free text allowing the mentor or mentee to describe themselves in a more deliberate fashion. Once a mentee has completed his profile, the algorithm presents the student in search for a mentor with the ten best matches. The first five choices are mentors with a broad specialty (e.g. internal medicine) concording with the mentee’s area of interest. If available, the first three matches actually match the specific specialty the mentee has selected in his profile (e.g. cardiology). The mentee can read the free texts of the proposed mentors and if inclined to, choose one as his mentor. To analyze the matching process and the resulting mentor-mentee-relationships we conducted an online evaluation at the end of the pilot semester in October 2008. The evaluation consisted of a questionnaire for mentors and mentees, respectively (n=87 for mentees and n=47 for mentors).

Results: 75.9% of mentees stated that the proposals by the matching algorithm met their expectations. While some mentees altered their profiles to get different mentor propositions, 88.6% of the students chose their mentor from the first ten proposals. 92.4% of the mentees judged the free text to be crucial for their final selection of their mentor, whereas 83.5% found the medical specialty to be the most important selection criterion. Interestingly, prior acquaintance with the mentor had an impact for 13.9% of mentees only. Importantly, 95.2% of the mentees kept their mentors over the pilot semester and 100.0% of the mentors found their mentees to be a good match.

Conclusions: We have developed an innovative and feasible method to match mentors and mentees by combining a profile-based automated algorithm with free choice for mentees. In the evaluation of the project, both mentors and mentees were highly satisfied with the selection of their partners.