What makes a good reference letter for a junior faculty position

Question

I'm in the process of applying to junior faculty positions (e.g. Assistant Professor). This will be my first time I ever apply for an academic position.

My direct supervisor asked me to write a first draft for my own reference letter for this application. I know that this can be considered unethical but I cannot change it. I have never before written such a letter nor read one. The examples I found online are not very helpful.

What are the elements of a good reference letter?

What distinguishes an excellent reference letter from a good one?

Answer 1

Disclaimer: I have not served on a faculty hiring committee, so this answer is not based on first-hand experience. What is expected of a recommendation letter also greatly depends on the field, country, and whether the department is teaching-focused or research-focused and how "elite" it is.

Academic recommendation letters are quite different from non-academic reference letters. A good academic recommendation letter should be detailed. Simply saying "spore234 was my student and they were the most smartest, and hardest working student ever" is not going to cut it. In particular, it should discuss your research, its significance, and what you contributed to each project. This is particularly important if your research is collaborative -- recommendation letters can elucidate what you contributed relative to your coauthors. It should also discuss teaching and service if applicable.

That is probably why your supervisor is asking you for a draft -- you are more familiar with your work than she/he is and can thus provide more detail. Hopefully your supervisor will heavily edit your draft (please encourage her/him to do so), but you can help provide details to include in it. Don't underestimate how little your supervisor may remember about what exactly you contributed to this or that project, or even what the project was. By drafting the letter, you are providing a reminder.

Furthermore, recommendation letters sometimes include a comparison to other students. E.g., "I have advised 10 PhD students during my career and spore234's research contributions would place her in the top two. Spore234 is comparable to my former student X, who has gone on to a successful career as a now-tenured faculty at the University of Y." This is still a subjective assessment, but it's a more precise way to express an opinion than simply spouting superlatives. (I would be hesitant to draft such a paragraph about myself, but perhaps you can let your supervisor fill in the blanks.)

Of course, a recommendation letter should also include the usual pleasantries, some detail about the recommender and how they know the subject, and some forward-looking discussion of future potential.

Answer 2

If this is in math I think you may have misunderstood your task. Math recommendation letters are weird. They start and end with a paragraph or two of a normal recommendation (comparisons, opinion) and in the middle there’s 1-2 pages of detailed description of the persons research. This will have a little bit of judgement (“I think this is their strongest paper” or “this one of the best papers in the subfield in the last five years”) but mostly about the papers not the candidate. This is kind of a problem because it means writing letters in math is more work, but a short letter from all but the most famous people will hurt the candidate (especially a short letter from an advisor).

What you should do is write the middle section with minimal judgement and let your advisor add the judgement. That is write a kind of annotated bibliography of your work, maybe grouped by research program starting with the work you think is the best. Include a little judgement comparing your own paper (“this is X’s most important work”). Then your advisor will add the first and last paragraphs and more judgement.

You can even ask your advisor (again, if it’s math) “I’d be more comfortable writing an annotated bibliography, would that be enough?”

At some level this even gets formalized. When I went up for tenure I was asked to write an annotated bibliography to be sent to my outside letter writers to decrease their workload and increase the number who will actually agree to write.

Answer 3

1. You can google and find plenty examples and tips: https://www.google.com/search?q=writing+recommendation+letter+faculty+position

2. In general, a good recommendation letter, regardless whether it is academic position or postdoctoral fellowship, should be supported by evidence as much as possible. Not just a hand-waving, that he is excellent and brilliant. But he is excellent because he published in such and such journals, trained such and such students, achieved such and such accomplishments etc etc. Try to support everything. Try to convince.

3. For academic position (your case) state explicitly why you fit for this position (e.g., your track record, mentoring / teaching experience).

4. Excellent vs. good is the difference of how enthusiastic the letter sounds. "He is one of the best in..." or "you will not find better candidate for this position" will likely make a letter excellent. But as I wrote in item 2, the letter should not be a mere hand-waving. But given that the letter does provide good factual information, the more enthusiastic the letter is the better. Put it simple: if the person who recommends you is of very high opinion of you and what he/she describes is supported by the facts (e.g., publication record, mentoring, teaching), then the letter will more likely be excellent.

Good luck!