One academic skill that many folks first experience as graduate students, which I suppose goes hand-in-hand with teaching for the first time, is writing letters of recommendation. Up until now, we’ve always been the letter seekers, asking teachers, professors, coaches, and others for their written support. But many of us reverse roles at some point in graduate school, often after having served as a TA or mentor for undergraduates. Now we become the voice of experience, having to evaluate a candidate for a job, graduate school, or some other type of program.
Of course, like many grad school skills, this one is not explicitly taught. Faculty at least can draw upon their experiences serving on committees that actually review such letters, but graduate students usually haven’t yet been in that position. How are we supposed to know the properties of a well-written letter?
Having written several letters myself now, I have developed a few habits that I hope are good. I always include certain details that I think will be important to the reader, such as who I am and how I know the student. I try to discuss my experiences with the student in a way that is relevant to the goal of the particular application (e.g., medical school, a summer research program, etc.), sticking to specific examples and realistic assessments — I’m sure evaluation committees tire of reading generalities and effusive but meaningless praise.
However, so far this has all been based on my intuition, not on any real knowledge or experience regarding what makes a good letter. I would be very interested to hear the perspectives of both other graduate students as well as faculty members on this issue, for the benefit of all who write and read letters of recommendation.