Summary – Seven Research Reports on M-P Matching

By Barry Sweeny, January 2011


Kessler, S. How to Start a Mentoring Program, Apr 6, 2010 , INC Magazine (Bisness mentoring)

As a researcher, I can tell you that how you best match people is probably the issue about which we know the least.” says Dr. Tammy Allen, a professor of psychology at the University of South Florida and co-author of Designing Workplace Mentoring Programs: An Evidence-Based Approach.”There are some companies that have created these algorithms that are used to match mentors and protégés almost like an eHarmony system for dating.” She’s also seen companies use random processes for matching, such as picking names out of a hat. But, she says the programs in which the participants have
some input are usually the most successful.”

One way of allowing input (is for) mentors and mentees to fill out an information sheet about what they hope to offer or obtain from a mentoring relationship. McGraw-Hill has a system for making mentoring matches that includes a questionnaire, phone interview, and committee recommendation for each mentor and mentee. Another way may be to offer the mentor or mentee several options for a partner and allow them to choose one.

Dr. Liz Selzer, a consultant with a California-based mentoring consultancy group, suggests matching people based on who they’ll get along with best. “If people get along, they’ll stay in the pair longer,” she says. “In case a pair doesn’t click, … set up a way for people to get out of the relationship and find another match without hurt feelings. Set up a “check-up” soon after the relationship begins.”

Editor’s Notes

1. When phrases like, “some companies”, “also seen”, “another way may be”, and “suggests” are used in a book specifically titled ‘Evidence-based’, we have to read these as ideas or options, and conclude that the author (a researcher) has found no evidence to recommend best practices. Note the first sentence in the next report as well.

2. There was no reference to relating strategies to program goals, so we cannot conclude that any ideas offered are better for any one purpose or kind of program.

Editor’s Summary of Kesler, S. (2010) Recommendations

1. Allow partners input into the match by completing a survey on what they offer and hope to gain.
2. Match for “getting along” factors
3. Do a check on the natch after a bit.

What makes for a good match? In “Using Telecommunications to Develop Mentoring Relationships”. Notes, the newsletter of the Center for Children and Technology, June 1996, Volume 4, Number 1 (Youth

The jury is still out on this one.” Experiments have been conducted with a variety of approaches to matching students with mentors: teachers have selected mentors for their students, students have selected their own mentors, and a handful of random matches have been made. These experiments have helped
to shed light on this issue and show that there are advantages to each approach.

When students selected their own mentors, students were often motivated to immediately begin their on-line relationships. These students tended to base selections of mentors on criteria such as: shared hobbies, career interests, or background information.

When teachers selected mentors for students, teachers appeared to consider a wider range of criteria: e.g. mentor’s tone as exhibited in their written descriptions of their work in relation to their students’ personalities.

Random-matching allowed for streamlining the process, but did not ensure that the students would have much in common with their mentors.

Editor’s Notes

When the author states there have been “experiments” but does not describe what that means, there are only two options: (1) different programs have tried different methods of matching, and (2) experimental research (treatmernt and control groups with random assignment). We have no way of knowing if it
is one or the other, so we can only assume the minimum (#1). Therefore, we are given no evidence that one is better for any particular goal.

Editor’s Summary of “Notes” (1996) Findings (not recommendations)

  • 1. Three methods have been tried in “experiments”
    • Persons who know the protege make the match
    • Proteges select their own mentors
    • Random matching
  • 2. Each approach has advantages (and we assume, each had limitations as well.)
    • Person who knows the protege make the match – wider range of factors considered
    • Proteges select own mentors – narrow, shared interest criteria, protege was motivated
    • Random matching – an easier, faster process with no known beneifts for proteges.

Alsever, J. (2009) How to Start a Mentorship Program,, Feb 06 2009 (Business mentoring)

Retrieved Dec. 14, 2010 at

Human resource executives at Nationwide Insurance found themselves “over-thinking” the matches according to discipline and personality. They’ve discontinued that approach and made it much more random, and it works better. “We were trying to engineer the relationships,” Plimmer says. “There is such a wide spectrum of things you are bringing to that relationship, and we were only looking at a couple of dimensions.”

Editor’s Note

Persons charged with making the mentor-protege matches originally thought they could design the best matches buy considering a list of and aligning common factors. (It does seem so logical to do so.) They apparently found that good matches and mentoring involved many more factors than could reasonably be analyzed and used, and that this method produced little benefit or reason to make the matching effort worthwhile. (Such a method typically produces a match in which the pair gets along well, and this is desired because it is perceived as necessary for the match to endure.) However, other more desired benefits for the business were not found to occur.

Editor’s Summary of Alsever, J. (2009) Implied Recommendations

1. Matching based on common factors of the people would logically seem to produce a match in which the pair gets along well and it would endure.
2. Matching based on common factors of the people produced no desired benefits for the business.

Larsen, P. (2007). “Student-to-Professional Mentoring as a Supplement to Public Relations Education” Paper presented at the annual meeting of the Association for Education in Journalism and Mass Communication, Washington, DC, Aug 08, 2007 (College student mentoring)

Retrieved online on Dec. 12, 2010 at

This study found no correlation between matching for perceived similarity of partner personalities and satisfaction with the effectiveness of the relationship.

Editor’s Summary of Larsen, P. (2007). Recommendations

1. There is no value for significant satisfaction with mentoring attained when pairs are matched for common personality factors.

Feiman-Nemser, S., & Parker, M.B. (1990). Making subject matter part of the conversation in learning to teach. Journal of Teacher Education, 41(3), 32-43. (Novice teacher mentoring)

The teacher induction programs that have succeeded have used mentor teachers with similar grade level and subject matter background. These master teachers have been invaluable in helping teachers develop management skill, subject matter confidence, and pedagogical knowledge

Editor’s Note

Study author Sharon Feiman-Nemser is one of the most respected researchers and has a research record extending over more than thirty years. Her conclusions are typically very valid and solidly based in evidence and excellent research practices.

Editor’s Summary of (Feiman-Nemser & Parker, 1990). Recommendations

1. Successful mentoring programs are so, at least in part, because they used matching for the working conditions most likely to help the mentor meet the needs of the protege.

2. When working condition factors are the basis of matching, mentors have been “invaluable” and the proteges have improved in the prerequisites for excellent performance.

Beans, B. E. (1999). Mentoring program helps young faculty feel at home. APA Monitor Online, 30(3). American Psychological Association Newsletter. (University faculty mentoring)

This study found that junior faculty received assistance and support from mentoring programs that paired them with senior faculty outside their own departments because it lessened the risk of revealing feelings of inadequacy to their peers.

Editor’s Notes

1. This is a study in one situation which uses one matching method that proteges seem to like. It is not an “impact” study, nor treatment – control group experimental method, so it does not demonstrate that the method protegs like was a better method than any other.

2. A low risk mentoring environement can be translated to mean that trust existed within the relationship.

3. A low risk, trusting mentoring environement can be quickly created when the mentor is not also the protege’s evaluator, and when the pair agree early to confidentiality in their discussions.

4. Matching outside one’s functional unit means the mentor may not have the experience to help the protege with discipline-specific needs – needs which are critical to successful completion of task assignments. However, the impact of this is moderated by the goals of the mentoring program.

5. This is why a more effective approach includes common working factors in matching, and uses othe methods (as in #3) to create the low risk environment for deeper sharing and growth.

Editor’s Summary of Beans, B. E. (1999) Implied Recommendations

1. Matching across functional units (depertments in this case) lowers the risks of relationships, and that low risk environment is valued.

2. A low risk mentoring environement is a prerequisite that is valued as it can quickly produce protege sharing on a deep level.

3. Deeper protege sharing is a prerequisite that is valued as it helps mentors discover (and target?) the real problems proteges have.

Favaretto A, Fanton E, Zampieron A. (2010). (“Student nurses’ evaluation of clinical tutors” – Article in Italian). Professioni Infermieristiche. Apr-Jun 2010; 63(2):93-8. (Health care mentoring)

The study gaol was to assess how nursing students perceived their tutors in the light of the personal and professional characteristics of both trainers and trainees. Results showed that the personal and professional characteristics of trainers and trainees are not very influential.

Editor’s Note

1. “Tutor” is the literal translation of the Italian text, but should be understood to mean mentoring and a larger support concept than tutoring typically suggests.

2. The terms “trainers and trainees” should be understood to mean mentors and proteges.

Editor’s Summary of Favaretto A, et. al. (2010) Recommendations

1. The personal and professional characteristics of the mentoring partners was not found to be a significant “influence” in creating mentoring effects or impacts.

The findings from the above research reports are combined and compared with each other and with another very comprehensive report, “Excerpts Related to M-P Matching” From – Sweeny, B.W. (2007) Leading the New Teacher Induction and Mentoring Program. Best Practice Resources, Wheaton, Illinois, USA”

Click the title to read that synthesis document.

See  a VISUAL representation summarizing the matching methods used in the referenced research.