For the first time, I've been asked to write a letter of recommendation for someone applying to graduate school. This person teched in my grad lab, where I supervised him on several projects and co-authored a paper with him.
I told him I'd be happy to write a letter, but pointed out that of course he'd be better off with a more senior person than me, if possible. He's got two PIs writing him letters, and he thinks I'm the best third letter. So here we go.
I've read some grad school letters of rec before, when I was a student member of the grad school admissions committee. As best as I can tell, here are the major components to the letter:
0. Who you are (presumably optional if you're a fancy big name scientist).
1. How well you know the person, and what your interaction has been.
2. Their scientific intellectual capacity.
3. The quality of their lab work.
4. Personality.
Any one of these topics can sometimes go awol, which suggests to the reader that the writer didn't have great things to say on the subject. I don't recall ever reading a letter that said anything less than stellar about the applicant, although there were more-emphatically-stellar and less-emphatically-stellar phrasings.
Am I missing anything? Do you tackle letters of rec in any particular way?
15 years ago
13 comments:
Seems like a pretty good formula to me. Handy too, as I've got to write some letters for students this week.
I think the formula would differ though if you'd only had the students in class, versus actually supervising in a research setting. For students I've just had in class I tend to talk about their academic achievements in my class, their ability to work independently/in groups/leadership abilities, writing, quantitative and oral communication skills.
Hi, although I'm not high up the ladder, I've been ask to write and have written two letters for others (so far) and for both I've asked a faculty member for advice. What I was told was practically what you say here , and what SW does (write only about what you know for sure about this person, and say why you come to this knowledge), - plus think about what this letter may (one day) tell about yourself. If you praise somebody who's not worth the praise, and they take this person on for grad school, they will surely find out whether the praise was real.
As a postdoc, I've written two such grad-school letters for techs I've trained. I approached it pretty much as you listed. The first time I wrote a letter, I showed it to my grad PI (the tech knew I was going to do this, and still asked me to write the letter). My PI said to not mention personality issues unless (a) they're a danger to society or (b) it relates directly to science/research [tenacious, team-player, prefers to work independently, etc.] He also said that he likes to see at least one letter from a less-senior writer (i.e. postdoc) because they generally work more closely with the applicant. As such, their letters are usually more pointed and contain more specific details that speak to the applicant's potential as a grad student. He felt this was enough of a benefit to outweigh the lack of breadth of experience postdocs would have to draw on (he is the BEST tech I've ever had, etc.) Good luck!
Part of my job is to write recc. letters for professional schools.
I agree with all the previous posters say, but specifics are important. For example, ...Student was asked to perform X procedure, and Student troubleshot and modified the procedure so that X procedure is now more efficient. The entire lab uses Students modified procedure to great success...
I disagree however about personality. You shouldn't spend a lot of time about how cheery Student is (or isn't) BUT you should talk about things like maturity, tenacity, ability to shift paradigm, ability to work well with teammates, etc.
I disagree however about personality. You shouldn't spend a lot of time about how cheery Student is (or isn't) BUT you should talk about things like maturity, tenacity, ability to shift paradigm, ability to work well with teammates, etc.
Exactly! In fact, I would consider it unprofessional to include opinions on the extent to which the person is outgoing, introverted, subdued, jovial, lighthearted, serious, etc.
Also, you should feel free at the end of the letter to give your judgment about the likelihood of the future scientific success of the person.
Yeah, I was mostly thinking of "personality" as "easy to work with" not "saturnine" or whatever. But thanks for the clarification; it's good to know where the line is.
Most of these comments are right on target. A few notes. I've read a lot of these letters now (I'm the chair of our admissions committee) and written my share.
Generally, your formula is great. The key is specifics. The people reading these have (usually) learned how to read between the lines. I would also point out that students can be stellar in different ways.
On the other hand, you should be honest about the qualities of the student. Saying a pretty good student is stellar is going to cost you in the long run. You can't say "this person's the best I've seen in 10 years" for two different people in the same year. (We once saw this on a fellowship study section. Neither person got their fellowship.)
Finally, simply agreeing to write the letter is a sort of endorsement of the student. (Remember, you can always tell them you can't write a stellar letter and give them the opportunity to ask someone else.) For many of these situations, at least one person on the committee will know who you are and will say "Dr. J and Mrs. H think this kid is great. We should accept them."
Finally, about writing a letter as a young professor about someone you knew when you were a grad student. This is totally fine (and often helpful) as long as (1) the person has other letters from outside that lab to help them [all three letters from the same lab is generally frowned upon] and (2) they have a reason for picking you [e.g. you are the person who was directly supervising them].
Hope that helps.
All good comments and suggestions.
I try to make my letters as individual as possible, bringing in very specific examples of outstanding work or instances in which the candidate overcame some major obstacle. I know I pay attention to such letters.
I would add, also, that potential supervisors are very interested in the personality of candidates. How well someone works with others, is a good role model, manages people well (supervisory position), etc., is important information, in addition to technical expertise.
A person may be technically qualified for a position, but is not a good fit for various reasons. For example, someone who does stellar work alone, but has an abrasive personality, would be perfect for some positions, but a disaster in others. As a supervisor, I need to have such information to make an informed selection. If I can't get this from the letters and have some concerns, I will call former supervisors/advisors and ask direct questions. If they don't give a direct, positive answer, I know there's a potential problem.
However, reference letter writers should be cautious in describing personal characteristics. Mentioning a candidate's leadership abilities, teamwork, etc. is appropriate. If you can't say something positive, however, then leave that aspect out or decline the request altogether.
For really outstanding students/ former staff, I always add a closing statement that if the opportunity arose, I would hire (or rehire) that person. This sends a clear message to the search committee that you have no reservations about that individual.
Best to be specific - describe instances where the person did something that demonstrates whatever good quality you're attributing to them. Otherwise the candidate sinks in a sea of generic adjectives.
Just one thing to add to this great discussion. These days candidates have to sound very good to even make the first cut. There's a general reference letter inflation to go along with grade inflation. It may be best to let the student know if you are not going to give him/her a great reference. In the past giving a balanced review was the right approach as it made the letter believable, but I'd be careful of that approach now.
And one more thing to add to my last comment. The status of the writer makes a difference. Dr. Famous writing a good but not great reference will always carry more weight than Junior Doctor's great reference so that would factor in as well.
I'm curious -- what's the take on my comment. Wish I had posted this before the tail-end of the discussion.
Aurora, I agree that in the US, at least, all applicants are above average, as it were. People mostly stay silent on topics that they think are not in the applicant's favor.
One more thing to add that nobody mentioned- be especially cautious if you're writing letters for women vs. men. Most of us have an unconscious bias toward discussing the personal traits or activities of women more than we do of men.
Case in point- women who have children often get letters that say "in addition to juggling a family" etc. Which is totally irrelevant and inappropriate, and yet most people mention it.
Beware of phraseology that reinforces gender schema (see this link for an example of a study that documents the prevalence of these kinds of problems).
Post a Comment