Getting Great Letters of Recommendation

Friday, February 2, 2001

We were three Ph.D. students from Stanford applying for the same academic position at a Canadian university, yet I was the only one who was asked to come for an interview. I wasn't the brightest, I didn't have the best grades, and I didn't finish my dissertation in the shortest time, but when it came to enthusiasm, quickness on my feet, and outright creativity, I knew I was tops among the three. However, none of these traits were reflected in my transcript. The only way interviewers were going to find out I had these qualities was through my letters of recommendation. It was those letters that got me -- and not my competition -- to the next stage.

Letters of recommendation are often the first independent assessment of your capabilities, performance, and potential that are seen by a search committee. It's important that they be first-rate. If there is to be any follow-up by telephone or in person, it will most likely come if your letters are from mentors who say great things about you.

Anything less just won't cut it in today's competitive job market in the sciences. The critical role of these references is captured nicely by Irene Peden, professor emerita of electrical engineering at the University of Washington, who says a good recommendation letter is enthusiastic and does not "damn the subject with faint praise or attempt to give a balanced view by articulating the candidate's shortcomings as well as strengths. There are appropriate places for that, but ... not in letters of recommendation."

What do search committees look for in letters of recommendation? They want
to know how you compare with others in your field of roughly the same experience -- i.e. your competition. What are your capabilities as a scientist? Do you show promise of continued development and professional growth? Do you have the potential to direct the work of others? What is your interest in, and hopefully experience with, teaching and working with students? Perhaps they have seen you address classes or seminars and can report on your style and effectiveness as a lecturer. Any information about your services to professional societies and journals would also be useful.

As Ms. Peden notes, you don't want to be damned with faint praise. Phrases such as "is reliable," "can be counted on to be in the lab every day," and "works well with other researchers" are OK but only if they are followed up with specific anecdotes that show how you really stand out.

All of your letters will have some things in common since they are all about you. This commonality is good, since it reinforces a positive image of your work. However, each letter should also be unique. Specific aspects of your education, character, and capabilities, as seen from the recommender's perspective, should be included.

So how, and from whom, do you obtain great letters of recommendation?

The key to getting such letters is to treat the process with forethought, not as an afterthought. You need to know your recommenders well enough over time so that they can say substantial things about you, backed up by firsthand experience and a reasonable amount of detail.

In principle, this sounds easy enough, since most of you will have been in the same working environment as your potential recommenders for a minimum of one, and more typically, three to four years. However, just being around a future letter writer is not enough. Recommenders may be asked to write a dozen or so letters each year, and you need to stand out in their minds as unique. That means making sure that over time these people become aware of your activities and accomplishments.

This is where forethought comes in. When you give presentations at university seminars or professional conferences, take note of who is in the audience. Follow up by sending these people copies of your talk. Take time to socialize with junior and senior colleagues at laboratory or department functions. Also, be sure to discuss your evolving career interests with junior and senior faculty members. Finally, don't be afraid
to ask senior scientists for advice about various career possibilities. The vast majority will welcome the opportunity to chat, (after all, they were once asking for the same guidance) and in the process will come to know you in a more well-rounded way.

How many letters of recommendation should you obtain? Three to five is usually the right number. Obviously your primary academic/dissertation adviser should write one of them. So should your supervisor if you are a postdoc. Someone who addresses your teaching interests and capabilities would be another. Also, consider obtaining a letter from someone in industry if you have had substantial interactions with the person during your graduate or postdoc experience.

What happens if your relationship with your primary adviser or postdoc supervisor is troubled in some way? This is a difficult question. Mary Morris Heiberger and Julia Miller Vick, who write this site's Career Talk column, addressed this problem briefly in a September 18, 1998, column. They recommend that you "ask the most senior person in your department whom you trust to write a letter for you that will, according to that person's best judgment, address the situation directly or indirectly."

I also suggest that you take the approach of asking different people to write about the different aspects of your work. In this way, important people with whom you have had some difficulty personally could be asked to write specifically about a success of yours in the laboratory, for example.

Another question that sometimes arises is whether to ask a "big name" scientist who may not know you well, versus someone of lower stature who is more familiar with your work. In almost all cases you should go with the people who know you best, who can demonstrate a personal knowledge of you gained through a long-term working relationship.

Keep in mind, however, that you don't always control who will be asked for their opinion about you. A senior, highly respected scientist in your department or laboratory, who may not even know you well, may be asked about you informally even if you did not list the person as a recommender. It doesn't hurt to make sure that such people are aware of your work through one or more of the approaches noted above.

The timing of your request for letters can be important. If you have just accepted a postdoc position, now is the time to get letters into your file from the people you knew as a graduate student. Even if they aren't used for a few years, they will capture a crucial period in your education. You want to have such letters put in your file while the memory of your recommenders is fresh. You can always go back to them for updates if appropriate. In some cases, you may want your recommenders to write two letters, one for a future academic position and one for a forthcoming postdoc or industry position. If you are completing a postdoc, then clearly your current supervisor will write a letter about your research experiences, but he or she could also talk about your academic interests.

At least a few months in advance of the need for such letters, you should sit down and have a thoughtful conversation with your potential recommenders about the kind of job you seek. Discuss the important aspects of your relationship as they relate to your application for an academic, postdoc, or industry position.

If you are seeking a professorship, talk about the desired balance between teaching and research, and graduate and undergraduate emphasis. If you are looking for a position in government or industry, talk about opportunities for both basic and applied research, and about possible publication limitations in a proprietary environment.

Don't just supply your letter writers with a copy of your C.V. Also provide one or two pages, perhaps with the main points in bulleted form, about things not in your C.V. that you wish to have expanded in the recommendation letters. Remind them of the particular way you approached and solved a problem, the initiatives you took with colleagues, and the feedback you received on your teaching evaluations.

Remember, your C.V. tells what you did. Your letters of recommendation tell how well you did it.

Strike the right balance in your approach. While you don't want to appear to tell your recommenders what to say or how to write letters, you do want to give them needed background (and reminders) about points they will want to write about anyway. Most recommenders appreciate this "assistance" if it is presented in the proper way.

Finally, be sure to write a formal thank-you note to your letter writers. Keep in touch with them as well. They are interested in the outcome of their efforts on your behalf, and no matter where you go next, they will continue to be your professional colleagues.