In mathematics in the USA, is it appropriate for your teaching letter of recommendation to be from a previous university?

Question

I am finishing a 3-year postdoc, and will be applying to many universities for positions, including teaching colleges and research universities. I have 4 good research letters, but none discuss my teaching. I had excellent teaching letters from my previous university where I received my PhD, but I am not as close to the professors in this university.

Would it be better to obtain a teaching letter of recommendation from my previous university that I know is very positive, or to obtain a potentially less positive letter from my current university?

Answer 1

I don't know whether it's better, but you may have more opportunities for a teaching letter from your current institution that you realize. Especially if your postdoc is in the US, there may be someone in the department (for instance, the deputy chair in charge of undergraduate teaching, whatever the position is called) who is supposed to help arrange the writing of such letters as part of their job. You could ask this person to sit in, or ask someone to sit in, on your class and write a letter based on that and department data about your teaching.

If you have excellent teaching letters from your PhD, you may even want to include that as well as a more cursory letter from your current school.

Answer 2

I think it depends on what kind of job you are looking for. For research universities, I suspect you can get away with it; I actually used a teaching letter based on recitation teaching from graduate school when applying for TT positions 4 years after graduating, because for a variety of reasons, there wasn't anyone else I felt comfortable asking for one from. Honestly, I don't think job search committees put much stock in teaching letters just based on classroom observation. Any value they had has been inflated away by the fact that they are always positive, and to be honest, most mathematicians at research universities don't worry very much about teaching skill when looking at hiring, beyond not wanting to hire someone notably incompetent. If they do, they're much more likely to try to judge from your interview or from student evaluations rather than a teaching letter.

If you're looking into more teaching focused schools, it's harder for me to say. Maybe I'll just leave that hanging and let someone else answer.

EDIT: To address the question of getting two letters, I agree with Noah. I want to be clear that this advice only applies to research universities, but I think getting two teaching letters completely gets wrong the risk/reward calculus for teaching letters. They cannot get you a job, they can only lose one for you. Probably they won't even look at teaching letters until they get to the short list, but if they do, it will be to sort out bad teachers, not to separate competent and good. By far the most important thing about a teaching letter is that it doesn't say anything bad. By getting two, you are doubling this (small) probability for absolutely zero benefit.

Answer 3

In your situation (where you had serious teaching experience in grad school), I would probably get a letter from your old writer, but send that person and updated list of what you've taught and your student evaluations from your postdoc. That way their letter can make it clear that your teaching continues to be strong.

I agree with Henry's point that you're probably overestimating how hard it is to get a letter from someone at your new place. However, only consider two teaching letters if and only if you're applying to a job with a strong teaching focus. Don't send two teaching letters for a research-focused job.