Recommendation letters

Will I write you a letter of recommendation? Writing student recommendations comes with the faculty job, and I usually write if asked. But since this academic territory is often so unfamiliar to grad and undergrad students, I thought it helpful to give some general guidelines and advice.

Strong letters usually come from long and close relationships with faculty. We take these letters seriously, since we often write these letters to our colleagues at the same pool of schools and employers. Our reputations are at stake. So students should request a letter with equal seriousness and care.

Good letters also take time to write, especially those for graduate school and (most of all) for the academic job market. Approach your professors as early as possible–ideally at least a month in advance. Faculty are extremely busy, and last-minute requests might get last-minute effort.

I think I’m similar to other faculty in that I generally write recommendation letters for students who:

  • took at least one course with me and excelled;
  • wrote a strong senior essay, thesis, or dissertation with me;
  • worked for me for at least four months; or
  • had some other close, personal interaction relevant to your application.

If none of these apply, it’s difficult to write a strong letter. Letters are usually explicit about how long we’ve known you, in what capacity, and how we rank you relative to your peers and our past students.

How can you find out if someone will write a strong letter for you? One way is to ask. Personally, I am frank with students, so as not to do them a disservice.

Once a professor has agreed to recommend you, three or four weeks before the due date you should email the following:

  • your latest CV;
  • your transcripts, GPAs, and (if relevant) GRE scores;
  • a list of where are the letters will be going; and
  • depending on where the letter is going, some supplementary information:

Grad school, scholarship, or academic job market applications: share your statement of purpose, research proposal, or other personal statements. Letter-writers take these seriously, and often discuss the specific strengths and weaknesses of the proposal in their letter. So try to ensure you letter writer sees a fairly polished version before the letter is written. This means you probably want to start discussing the proposal with your advisors at least six to eight weeks before the letters are due.

Job or fellowship applications: include a description of the job or fellowship, and what they are looking for in a letter. Often a job or scholarship application will offer specific guidelines. If they don’t, provide a quick description to your letter writers.

Try to send everything in one package, well organized. We letter writers are absent minded, and will lose the multiple pieces if you do not collate them for us.

Most grad school applications are online nowadays. This is less convenient than it sounds, since every school has a different custom form. Please fill out the maximum amount of information about your recommender that you can (including contact info) so that your recommender doesn’t have to complete the same info 13 times. This process takes a faculty member hours and hours.

For grad school applications, you want at least one to two academic letters. Two is probably better, with one from an employer if you have been out of school for long. So maintain contact with your advisors after you leave university.

If your application requires three letters, you might be tempted to send four. This is usually a bad idea: every additional letter after your best one lowers the average quality of your application. This is simple arithmetic. A fourth letter only makes sense if it will be glowing and adds a completely different perspective.

Finally, remember that (as academic applications go) letters from senior faculty usually carry more weight. Junior faculty who are poorly known outside their subfield (i.e. people like me) carry less. Strong letters matter most, though, and you may be able to build closer relationships with junior faculty less inundated with students. Keep this trade-off in mind. Probably a balance of junior and senior faculty recommenders is best.

Comments, criticisms and questions are welcome below.