Recommendations for letters of recommendation?

December 11, 2007 10:45 PM

What letters of recommendation should I get when applying for a Masters in Computer Science?

I completed my undergrad degree 7 years ago in Computer Science. Since then, I have worked at 2 different places of employment (software development work). I now want to pursue a Masters degree in Computer Science.

Each school that I want to apply to requires at least 3 letters of recommendation. I can fulfill this requirement by getting 3 letters from the 2 employers I have worked for since college. Should I also contact my undergrad professors who may or may not remember me and/or my work for additional letters? Would these letters from past professors even be applicable since it’s been 7 years since I studied under them? I haven't had contact with my professors since graduation.

posted by escher to Education (7 answers total)

I would think that the most relevant letters of recommendation would come from your most recent place of employment (if they are willing to do so). Undergrad professors won't carry as much weight the farther they are from your present-day experience.
posted by amyms at 11:00 PM on December 11, 2007

Don't be shy about getting recommendations from your professors, even if you think they won't remember you; it is part of their job. If you went to a larger school or had any professors who publish and go to conferences frequently, then your recommendation will represent a known quality to other schools. The community is not very big, especially within subfields, and chances are that the people who will be reading your application will know who wrote your recommendation and their caliber. Your recommender will serve as a yardstick by which grad schools will measure you.

So...
1. Choose a professor with a good reputation whose classes you have taken (caveat: the best choice will work in a subfield you want to study)
2. Write an email asking for a recommendation; it will help if you remind him/her what classes you took and what grade you got and anything special that might jog their memory
3. Send gentle reminders so they don't forget
4. Pray that they write something a) nice and b) personal.

If your work recommendations are glowing and your undergrad grades are solid, you won't even have to worry much about #4.
posted by Alison at 11:10 PM on December 11, 2007

Definitely get at least one from former professors. Even if they dn't remember you well, or at all, they can say "escher took my course CompSci 476: Coding and Foucault, earning a mark of A, based on 89%, placing 2nd in a class of 23". That's the kind of thing that Universities are interested in and which employers can't provide.
posted by Rumple at 11:59 PM on December 11, 2007

I am a computer science professor who writes and reads such letters. Different professors have different approaches, but I would say that in my experience unless you have reason to believe that the professor you ask to write a letter is likely to remember exactly who you are on the basis of some unusual or exceptional work you did as an undergraduate, then don't bother to ask them for letters.

The letters are really a small part of your application. Your transcript (where you went, what classes you took, and how you did) is generally most important, followed by your statement (to see if you are a match for the department's program), and your GRE scores (to do some sanity checking against the rest of your record). After that, your letters get a quick scan to see if there is anything notable in them. The letters are most useful when they highlight things that are not available from reading the rest of your record.

Most letters are pretty generic. They say "ecsher took my class and did well. He worked hard, etc". The longer it has been since the professor has seen you, then more generic they will be, and less likely to contain any information that isn't on your transcript. We already know what classes you took, from whom, and how you did. While it is true about what Alison says about subfields and networking, it is also easy to pick up on when the professor writing the letter is just writing a generic one versus when a professor knows you well and makes specific comments about your work and aptitude in that area. The former doesn't get much weight while the latter gets noticed.

If you did do some research work for a professor, and it went well, then by all means get them to write you a letter. But if you were just a student in a class, and you have someone at work who can make concrete comments on your motivation for grad school and skills, that would be far better.
posted by procrastination at 6:01 AM on December 12, 2007 [2 favorites]

I would imagine it also depends on the kind of Masters program you're signing up for. If it's a course-based one, then they are likely to be more concerned with your grades. If it's primarily research-based, then recommendations from people you've done research with (if any) would be very useful, whether they be from work or from the university.
posted by destrius at 7:36 AM on December 12, 2007

I used my undergrad advisor, my first employer and my boss at the second employer.

Honestly, I don't think it even matters. Nobody ever gets BAD letters of reference. So unless you're getting them from someone famous or important in the field (like, say, you got Bill Gates to write you a letter), it's just a checkbox they check--"yep, three letters." Things like your GRE scores and your essay are much more important. So just get three from whoever liked you enough to write it up and print it three times.
posted by jeffamaphone at 10:23 AM on December 12, 2007

Let me just say that, yes indeed, some people do get bad letters of recommendation. Personally, I just won't write letters if I only have bad things to say, but I have seen a few that were surprisingly bad. Some even said, essentially, "I told the person I had nothing to say, but they insisted that I write a letter, so let me tell you all their other foibles". The moral: if someone doesn't want to write you a letter, don't make them.
posted by procrastination at 6:54 PM on December 12, 2007