Reference letters for undergraduates

1. Policies

I normally agree to write a reference letter for any undergraduate who was a student in any one of my classes, or for whom I am otherwise qualified to act as a reference.
I may however refuse if I am really overwhelmed by requests, and if I feel that I cannot write a very strong letter.

When I write a reference letter, my goal is to write the strongest possible recommendation that I can, subject to the following constraints:

it has to be consistent with reality. In particular, if I only know you as a student in a class that I taught, and if your performace was not very good, then I can normally write a reference letter, but not an extremely positive reference letter.
If you were a student in a class that I taught, a reference letter I write for you will normally include rather detailed and accurate information ^[*] about your performance in the class, such as "his/her overall course mark was in the top ... %"
If I am writing many reference letters, which is often the case, then this places constraints on the amount of time I can spend on each letter.

I reserve the right to limit the number of graduate programs to which I will upload a reference letter. In most cases it is simply not reasonable for a student to apply to more than about 10 graduate programs, and I do not think that professors have any obligation to facilitate this ureasonable behavior.

2. What you need to do

I will ask you to send me by email

a copy of your transcript
a copy of your resume
a sample of a personal statement for a graduate program to which you are applying (if it's a reference letter for grad school).
a list of all the graduate programs you are applying to, together with all the deadlines for letters of reference (if it's a reference letter for grad school).

I will also ask you to make an appointment to meet me (online, for 2020-21). This meeting may take as little as 5 minutes or as much as a half hour. The motivation is that I think that reference letters are more persuasive if they give the impression that the letter-writer actually knows the student. After talking to you briefly, it is easier for me to convey this impression (whether accurate or not). Reference letters are also more persuasive if they contain some interesting or memorable detail that makes the letter stand out from the hundreds of other letters that a committee might read. So if I can learn something interesting or unexpected in talking to you, all the better. Of course, it is also fine if there is absolutely nothing interesting about you.

I am often available for appointments at the following times:

Tuesdays 2-5pm
Wednesdays 2-5pm
Thursdays 11am-noon

If you are in a very diferent time zone, I can also be available rather early in the morning (in Toronto) on some days of the week, eg Tuesday, Thursday, Friday. To make an appointment, please send me several possible times, if possible within the above periods, when you are able to meet.

If i meet you, I may take the opportunity to ask you about your experience as a student at the U of T, for example, just because I am curious about what life is like for our students and I have few opportunities to find out.

Please complete all of the above (the appointment, plus the emailed information) at least 2 weeks before the earliest deadline for any reference letter.

3. A few aditional remarks

If you were a student in MAT237 and you are now applying for a graduate program in a field related to mathematics, then I recommemnd that you look for letter-writers who know you from more advanced classes. This will give your applications the best chance of success.

If you were a student in a class I taught, then I will write a reference letter for you whether or not you found the class interesting. The quality of the letter also will not be affected by anything you tell me, for example, about how much you liked the class. In fact it makes me a little uncomfortable if someone praises/flatters me at the same time that they are asking for a favor. If you liked the class and feel a need to tell me about this, please find some other time to do so.

This page is dedicated to the memory of Scott McMillin. When I applied to graduate school, for various reasons I asked for and received a reference letter from the professor in an English literature class that I had taken as an undergraduate, Scott McMillin. He was a good teacher, slightly eccentric. I half kept in touch with my other letter-writers, both mathematicians, enough that they knew a little about what had become of me, but I forgot about Professor McMillin. Some years ago, when I was writing a lot of reference letters, it occurred to me that I should send him a brief email to thank him for his help all those years earlier. So I looked him up and found that he had died a couple of years before. Then some time later I thought that perhaps I would get in touch with his widow instead. But she had died in the meantime. Now (January 2020), a few more years later, I have decided to put the thanks here.

^[*] One reason I do this is that, in submitting recommendations, faculty members are typically asked to check boxes on an online form, indicating whether the student is in the
top 2%
top 5%
top 10%
top 20%
top 50%
by various criteria, such as research potential, communication skills, etc. I am told that more or less everyone lies all the time in doing this, so that if a student is in the top 15% of MAT357 (very good!) and I therefore click "top 20%" in all the boxes, then the student will be rejected by all grad programs. So my policy is to allow myself a little dishonesty in clicking these boxes -- what I judge to be the standard level of dishonesty, so that adjusted for universal exaggeration, it will give an accurate impression of the student's abilities -- but, because I don't like to lie, I also give honest figures in the letter, where I can put them in context.