Letters of recommendation: A department affair?

A recently posted query from a reader worried me. The query was about how to improve letters of recommendation, but since I was assuming the writer was a women, I worried that the problem originated in some sort of gender-related bias.

Whether or not it did, recent research lends weight to the idea that letters of reference do show a kind of gender influence that is undesirable. I’ll mention details of it below, but the question now is, what to do about it?

Perhaps it is a bit late in the year to do much now, but we probably do not have the means to accomplish a lot. So let’s suppose we do some more research and solidy identify some problem areas (e.g., men are seen more as active agents, their praise words are higher and more definite, etc), is this the sort of thing that can be effectively addressed in a department meeting? Are we going to find that there are people sitting there unmoved because they think that men just are more motivated, have more initiative and are generally brighter?

Another possibility would be to get some departments to agree to open their dossier files for a research project, and to convince some members of a social science department that analyzing them would make an interesting project. There are two women in philosophy who have been or are now PIs on NSF Advance grants, and they might help identify universities where this has some likelihood of succeeding.

Another thing would be for this blog to do a useful page on how professors can improve their letters for students, with a list of the common problems seen. (Actually, if we haven’t done it already, I’ll be happy to do that.)

Any more ideas?

One recent bit of relevant researchThe paper is online; here is the abstract:

Gender and Letters of Recommendation for Academia: Agentic and Communal Differences

In 2 studies that draw from the social role theory of sex differences (A. H. Eagly, W. Wood, & A. B. Diekman, 2000), the authors investigated differences in agentic and communal characteristics in letters of recommendation for men and women for academic positions and whether such differences influenced selection decisions in academia. The results supported the hypotheses, indicating (a) that women were described as more communal and less agentic than men (Study 1) and (b) that communal characteristics ave a negative relationship with hiring decisions in academia that are based on letters of recommendation (Study 2). Such results are particularly important because letters of recommendation continue to be heavily weighted and commonly used selection tools (R. D. Arvey & T. E. Campion, 1982; R. M.Guion, 1998), particularly in academia (E. P. Sheehan, T. M. McDevitt, & H. C. Ross, 1998).


Juan M. Madera, University of Houston

Michelle R. Hebl and Randi C. Martin, Rice University


I just discovered a post by Monkey which covers the research in the article, and she links to some of the background info from Rice University.

7 thoughts on “Letters of recommendation: A department affair?

  1. This is one of those areas where we’d do well working with other disciplines because I don’t think the issue of gender and recommendation letters is ours alone. The worst examples I’ve seen come from the sciences, in fact.

  2. Jamie, here’s a common explanation:

    Social cognition theory perspective that views people as self-organizing, proactive, self-reflecting and self-regulating, not just as reactive organisms shaped by environmental forces or driven by inner impulses.

    I’m not sure we need to refer to some theory here; ‘active,’ in contrast to ‘passive,’ might do.

  3. Well, what I meant was that it seems people are most defensive at the local level, whether that’s department or discipline. A study of letters of recommendation needn’t focus just on Philosophy to be helpful. In fact, I think that might be unhelpful. And I find, at least at my institution, the sciences seem better at working on some parts of the gender-gap problem. So studies of letters/workshops on writing better letters etc would be better at done across disciplines, I thought.

  4. It seems that there are two connected problems here: 1) how we think about and evaluate student accomplishments, and 2) how we communicate these accomplishments to others.

    I recently noticed that I was using similarly gendered language when evaluating my students’ class participation. I then went back and relied much more heavily on my notes re: the students’ particular comments during each class (the class was structured so that I happened to have a record of this), rather than factoring in my overall impression of the students’ participation; this made a noticeable difference in the grades. This seems to fit with the study’s findings that evaluators use gendered language to describe student accomplishments and then hold one category of accomplishment in higher esteem.

    My experience suggests that doing things like having a set criteria for something like ‘class participation’ and then relying on notes taken in the moment *might* help remove some biases – but I am skeptical of this method. If my overall impressions of student contributions were biased, what’s to say that my evaluation of each individual contribution was not biased as well? I guess it’s an improvement since it seems that my in-the-moment impressions were *less* biased than my overall impressions, but that is hardly a ringing endorsement of this method. Similarly, it seems likely that things like having detailed accounts of the student’s accomplishments won’t significantly help the problem when writing letters of recommendations.

    As we learn more about implicit biases, it’s clear that merely having the intention not to be a biased evaluator or letter writer isn’t particularly helpful. We know that methods such as blind grading can make a significant difference in removing implicit biases – but for things like evaluating class participation or writing letters of recommendation this just isn’t an option. So what can we do?

  5. cmh, it turns out that there is a way of having an intention not to be biased that can be helpful in some situations. The intentions have been labeled “inplementation intentions” and they have the structure of (something like) “If P, then I’ll A.” E.g., “If the next candidate is a person of color, I’ll ignore their color.” This works in some contexts; I haven’t seen any investigation of its role in grading papers, however.

    Frog, that’s a great idea. NSF Advance will have something about letter writing, I should think. In fact, the paper I referred to was written at a university with an nsf advance grant. And that’s about three miles away, and one of the authors has since moved to my university. (Putting 2 and 2 together, slowly…)

Comments are closed.