I'm planning on applying this fall to several statistics/biostatistics grad programs (probably Master's, maybe PhD; still deciding) and I'm trying to get the best recommendation letters I can.

For context, I graduated a year ago with a BS in Math, a BA in music, and a minor in Stats. I've been working in Pharma, though not in a position where I'm doing much math. I have one recommendation locked down, this being my Faculty Advisor for an REU I was part of and who I've kept up contact with. My other options are a bit dicier from there:

Option 1: My discrete math / topology professor from my sophomore and junior year. I got an A and B in these classes respectively. I went to office hours frequently and had a lot of good conversations and a generally good relationship with this professor. He wrote me the recommendation letter for the REU and I almost did research under him. That being said I haven't talked to him in over 2 years.

Option 2: My machine learning professor from my senior year. Got an A in his class, went to office hours frequently and talked to him about my interests. I asked him if he'd be willing to write a recommendation letter when I thought I was going to go to grad school sooner and he said yes. I've talked to him a bit over email since graduation but that conversation sort of petered out.

Option 3: My music professor from undergrad. Not at all math related but he taught me all throughout undergrad and we have an excellent relationship, still frequently in touch etc. I've gotten the impression most STEM departments won't care much about a recommendation from someone not field-related, but I know he'd write a great letter.

Option 4: My current work supervisor. I think she'd write a really good recommendation, and pharma is certainly biostats related, but we're completely on the manufacturing/engineering side (validation/compliance) and not at all on the clinical side.

I'll probably ask them all, but I'm wondering what you think the best bet is. For all cases, I'm planning on sending them a packet of all the things they might need to write the letter.

  • 4
    I think option 3 is a non-starter. Music is simply too far from biostats for your letter to have any reasonable value. Commented May 12 at 23:43

1 Answer 1


Recommendation letters should strengthen one's case for the position in question. For graduate school application this mainly means showing that the one has the background necessary for the chosen program, demonstrating the one's scholarly achievements, ability for hard work, and potential for doing research. In this sense, a recommendation from the professor who taught a recent course in probability, statistics or a related subject is probably the most important one. A recommendation from a professor of a mathy (but not necessarily statistics-related) subject is pretty close, especially, if they can present you as an excellent student. This is probably as far as people not in the field goes in most cases.

A music professor can characterize you as a hard-working person, and perhaps (alongside other recommendations) may demonstrate your widely ranging abilities and capacity for adaptation... but such a recommendation is clearly at a disadvantage compared to somebody closer to statistics/math. In other words, I would keep this at best as the third recommendation letter or even forgo this recommendation altogether (there is even a risk that such a recommendation does more damage than good - this may make application look as not serious or not capable of showing more relevant achievements.) It may be emotionally difficult to let go something, to which you have devoted much time and effort (which is likely the case with music), but it is just not relevant to your professional choice.

Not the answer you're looking for? Browse other questions tagged .