Writing a Letter of Recommendation

WRITING A LETTER OF RECOMMENDATION As a beginning independent investigator, chances are you will soon have to write a letter of recommendation on behalf of a student, a postdoc, or even a colleague. Your job as letter writer will be to describe the candidate’s strengths and weaknesses as they relate to the position or program in a way that is both thoughtful and personal. A letter that falls short of this goal will be of little value to those evaluating applications and will not help the candidate get what he or she is after. So, it pays to put in the necessary effort and time to write a “good” letter. This chapter provides insights and advice from experienced investigators on how to do so. It is not meant to be prescriptive but rather to offer some suggestions from which you can pick and choose. BEING ASKED TO WRITE A LETTER Letters of recommendation are ubiquitous in an academic research career. If you teach one or more senior-level undergraduate courses or have undergraduate students in your laboratory, you might have to write dozens of letters a year as students become graduate-school bound or look for employment. If you do not teach undergraduates and have primarily graduate students and postdocs in your lab, you will have significantly fewer letters to write—maybe for only one or two people each year. In addition to the people in your own lab, graduate students and postdocs in your department may ask you to write letters for them when they apply for fellowships or seek new positions. Colleagues might also ask you to write letters of recommendation on their behalf for various promotions or awards, but that may not happen too often until you are more established. For Whom Should You Write? As a mentor, you have an obligation to support students and postdocs in your lab in their job search and to help them find a good match for their abilities and aspirations. If they ask you to write a recommendation letter, it is customary to support them in this way. The best thing to do is to sit down with them and discuss their plans before they start applying for jobs. If their career goals are unrealistic, talk about what they need to do to become more competitive or help steer them Making the Right Moves A Practical Guide to Scientific Management in a different direction. If you do end up writing a letter for someone in your lab for a job you don’t think is a good fit, there are ways to put a positive spin on the negative (see page 6, “Writing a ‘Not-So-Enthusiastic’ Letter”). If people who are not part of your lab ask you to write a letter for them, it is up to you to decide whether to do so. Reasons to Turn Someone Down You should write a letter of recommendation only if you can honestly write a supportive letter for someone for a given position. After all, a letter of recommendation is supposed to be a tool for helping people obtain what they are after. If you don’t know the candidate well enough to write a good letter, let the person know. He or she will probably ask someone else. If you do know the candidate well but have some reservations, let the person know about your concerns, and leave it up to him or her to decide whether you should still write the letter. PREPARATION Read Some Sample Letters If you have never written a letter of recommendation before, read a stack of reference letters to see what works and what doesn’t. Most faculty have access to graduate student applications and the letters submitted. Junior faculty can also ask their more experienced colleagues to share “sample” letters they have written or have received from applicants (blocking out names and other personal information). The resource section of this chapter provides links to sample letters. Collect Information Once you know what a letter of recommendation looks like, make sure you find out as much as possible about the candidate for whom you are writing. One of the worst things you can do is to write a letter that is too generic. Ask for the following information: T I P : When reading a sample recommendation letter, try to put yourself in the position of whoever is going to read it, keeping in mind that you could be reading the letter late at night or when you are pressed for time. Ask yourself which letters stand out and then use them as models for your own letters. ʌ For people in your lab—you prob- ably already have a current curriculum vitae (CV), but also ask for as complete a description as possible of the position or program for which the person is applying. Writing a Letter of Recommendation ʌ For people who are not in your lab—ask for a current CV that includes the dates and locations of their training and a list of publications as well as copies of any papers that are in press or that have been submitted. ʌ For people who were in your lab but have since moved on—ask for a written summary of their current work, future plans, and why they are interested in the position for which you are writing the letter. ʌ For students who are not in your lab or department but who were in one of the classes you taught—ask for transcripts and check your class records. Many reference writers find it helpful to meet with the person who has requested the letter of recommendation. In such a meeting you can learn about the person’s long-term career goals and how his or her current research and other activities relate to these goals. Ask about the programs or positions for which letters are being requested and what the individual’s preferT I P : Don’t ask the candidate to write a draft of ence is and why. You should also discuss the letter for you. Most heads of labs say this rarely the candidate’s strengths and weaknesses saves time and sometimes leads to a weaker letter. as they relate to these positions. It is better for the candidate to provide you with all the necessary information, from which you can then pick and choose as you write your letter. If you cannot have a face-to-face meeting, ask the candidate for a detailed written description of these issues. Allot Sufficient Time to Write and Send the Letter Once you have gathered the data and marked the deadline on your calendar, make sure you set aside enough time to write the letter. It can take one to two hours, sometimes longer, to craft a well-written letter for one individual. (It takes less time to tweak this initial letter for different programs or positions the same individual has applied to.) The candidate should give you clear instructions on how and where to send the letter. Some letters have to be submitted online, some e-mailed, and others sent by snail mail. WHAT THE LETTER SHOULD SAY Keep the following pointers in mind when you craft your letter. Introduce Yourself and the Candidate Begin the letter by describing how you know the candidate. Briefly state your own qualifications so that people who don’t know you can decide whether to Making the Right Moves A Practical Guide to Scientific Management trust your judgment. But don’t go overboard and make the letter about yourself! For example, “During my years of training as a postdoc and now as a faculty member, I have worked with and personally trained more than xx graduate students in laboratories at the University of Michigan, Harvard Medical School, and, currently, at Emory University. In my opinion, candidate x is among the top 5 percent of the graduate students I have known.” (You may want to include students you trained as a postdoc, if you have not yet had many students in your own lab.) Present the Candidate Truthfully but Positively Devote several paragraphs to discussing the candidate’s scientific work and personality, following the suggestions below. Be specific. Give meaningful examples of achievements and provide stories or anecdotes that illustrate the candidate’s strengths. Don’t just praise by using generalities (such as “quick learner”), but say what the candidate did to give you that impression. These details will show you have a strong relationship and also bring the candidate alive on the page. If the candidate works in your laboratory, make sure the letter talks about his or her specific contributions to the lab’s research. ʌ “Student x is the first person to successfully master in vivo imaging in my laboratory and did all the trouble-shooting calibration and testing completely independently.” ʌ “Student x is the first student in my lab to have a first-author Cell paper. She contributed data for five of the figures and made an important intellectual contribution to the experimental design and to writing the discussion.” ʌ “Candidate x’s research was submitted as an abstract and was accepted for a podium presentation at the American Association for Cancer Research national meeting earlier this year. Although she is an undergraduate student, I treated candidate x in the same manner T I P : Don’t make the letter too short, because I would treat an experienced graduate it will give the reader a negative impression of the student by having her prepare and present candidate. Letters of recommendation should be the talk independently. At the meeting, between one and two pages. she presented the work and answered all the questions from the audience in a clear and professional manner. After the talk several scientists came up to express how impressed they were with the presentation.” Writing a Letter of Recommendation You don’t just have to describe the candidate as he or she is right now—you can discuss the development the person has undergone. ʌ “When candidate x came to the lab he showed promise but was not able to plan experiments effectively. Through his own hard work he has acquired this skill and has become an excellent scientist.” Quantify and compare. Find a way to quantify the candidate’s abilities, especially with respect to other scientists who have achieved success in the field and who the letter reader might know. Many letter writers rank the applicants according to their own measure of what makes a good researcher, graduate student, or technician. Depending on the position, most people rank candidates according to a combination of research strength, leadership skills, writing ability, oral communication, teaching ability, and collegiality. ʌ “There are currently 17 graduate students in our department and I rank candidate x as 4/17. Her bench skills are the best I have ever had in my own lab, but she still needs to develop better communication skills.” Address the requirements of the position. Make sure the letter discusses the most important skills and personality traits needed for success in the job. ʌ “I understand that you are looking for a person who will be able to establish a program to study nuclear structure and architecture using imaging technologies. Candidate x has performed similar studies using fluorescent imaging in living cells. Her mastery of these experimental techniques is demonstrated by her recent first-author publications in Journal of Cell Biology and Nature Cell Biology.” If you are writing a letter for someone applying for a position as a physicianscientist, you should comment on his or her bedside manner. For example, if you are a chief of service, comment on the candidate’s clinical skills and personality and how he or she interacts with patients. Talk about personal attributes. Does the candidate work well with others? Has she or he taught other people in the lab? Has the candidate shown motivation We look for more than research accomplishments—comments on creativity, originality, independence, and leadership are usually not overhyped and mentioned specifically in letters of the very best candidates. —Suzanne Pfeffer, Stanford University School of Medicine Making the Right Moves A Practical Guide to Scientific Management and commitment in her or his work? Often these descriptions are what distinguish a good letter of recommendation from a great one. Make it memorable. Put something in the letter that the reader will remember, such as an unusual anecdote, or use an unusual term to describe the candidate. This will help the application stand out from all the others. Beware of what you leave out. Remember that what is not said in a reference letter can be just as important as what is said. If you don’t mention a candidate’s leadership skills or his or her ability to work well with others, for example, the letter reader will wonder why. T I P : %½REPWXEXIQIRXWYQQEVM^MRK]SYVIRthusiasm for the candidate is often very useful in focusing the reader’s attention on your conclusions and your excitement for candidate. Express Your Willingness to Help Further Conclude the letter by offering to be contacted should the reader need more information or have questions. Make sure to provide your phone number and perhaps the best times to reach you. WRITING A “NOT-SO-ENTHUSIASTIC” LETTER If you write a letter for someone about whom you have some concerns—for example, the person lacks strong leadership skills or isn’t highly motivated—you can still write a generally positive letter, leaving out a mention of these qualities. This omission will send a signal to readers who will be looking for these details. A short letter will also signal that you are not overly enthusiastic about the person for this particular position. If the reviewers are interested in the candidate, they may call you for more details and you will then have the opportunity to explain your reservations. You can also mention negative things by putting a positive spin on them. For example, if someone has not published much you can say “candidate x has taken a long time to get experiments going but now has several papers in press and I think they will make an important contribution T I P : If the candidate is a good scientist but betto the field.” ter suited, say, for a faculty position at some schools and not others, some researchers handwrite that perception on top of the standard letter to those “other” schools. This way they still help the candidate obtain a faculty job without harming their reputation for good judgment. Yet another way to make a criticism sound less damaging is to say something like “when candidate x came to the lab his writing skills were poorly developed, but he has worked hard to improve them and has made great strides in that direction.” Writing a Letter of Recommendation Terms like “room for improvement,” “has worked hard to,” and “made great progress in” are useful and help turn a negative into a positive. WHAT THE LETTER SHOULD NOT SAY Avoid Irrelevant Information Letters should not mention anything that is irrelevant to doing the job, such as ethnicity, age, hobbies, marital status, and so on. A writer might want to say things like “candidate x understands how important it is to have a lab team that functions well and has organized several social events for the lab that have helped make it both productive and happy” to show that the person is social and easy to get along with. That is something that might be important to the job. In some cases, personal information may illustrate the ability to persevere and overcome adversity, qualities that are helpful in research. But statements like “candidate x is very religious and has demonstrated religious fervor uncommon for someone his age” has no bearing on whether someone will do well in graduate school or in a job. Personal information may be revealed if it helps explain a weakness in someone’s application. For example, sometimes there are acceptable circumstances for a gap in someone’s publication record—perhaps a medical condition or a family situation kept that person out of the lab for a period of time. It is okay to explain that in the letter of recommendation but only after discussing it with the candidate. Another example of a personal situation that might warrant discussion would be a case in which a postdoc is not able to ask his or her advisor for a letter of recommendation because of a bad relationship. If you, as the letter writer, know about this situation, you might want to mention in the letter that “there was a personality conflict but it does not reflect on the ability of the candidate to do the job.” Don’t Say Anything That Is Not True Obviously, all information in a letter of recommendation should be, to the best of your knowledge, accurate. But sometimes letter writers stretch the truth to make a candidate sound better than he or she really is, thinking it is helpful. This strategy can backfire. According to a member of a recent search committee, the head of one lab wrote letters for three applicants from his lab, and in each letter he stated that the applicant was “the best postdoc ever in my lab.” Most researchers say that the most effective letters they receive include some weakness to balance the strengths. Making the Right Moves A Practical Guide to Scientific Management You don’t help the candidate by overhyping them.You should write strongly and specifically about someone who is really excellent (say how and why they are special). But it’s okay to write a balanced letter, even for the top folks. —Suzanne Pfeffer, Stanford University School of Medicine BE AWARE OF GENDER BIAS Choose Your Words Wisely Two studies have shown that there may be some differences in the choice of words used in letters for male and female scientists. Both Trix and Psenka (2003) and Schmader, Whitehead, and Wysocki (2007) found that letters for women did not contain as many “standout” adjectives—words like “superb,” “outstanding,” and “excellent”—as did letters for men. In addition, Trix and Psenka found that letters for women had more doubt-raising statements, including negative or unexplained comments. If you say that Chris works well in a team—that might be interpreted differently depending on whether Chris is male or female. Exactly the same language can convey different messages because of people’s antecedent beliefs about gender differences. The reader might infer that Chris the man is a good team leader or someone who can work well with a team even when he isn’t a leader of it. Chris the woman, in contrast, may look as if she can’t or doesn’t lead. For both men and women, then, it is good to be explicit that the person is both a good team leader and can work collaboratively.” —Virginia Valian, Hunter College Test for Possible Bias Although it is not clear what the effect of the choice of words is, it seems plausible that the words you use will affect the reader’s perception of the candidate— the stronger the words, the stronger the candidate. In an exercise conducted at Writing a Letter of Recommendation the 2005 BWF-HHMI Course in Laboratory Management, course participants were asked to read two letters written by the same professor for two postdocs in his lab, one male and one female (see appendix), with similar qualifications and achievements. The professor used more T I P : Write a letter of recommendation for any specific and descriptive language, as well candidate, male or female, and then switch all the as more standout adjectives, for the male pronouns to the opposite gender. Read the letter candidate than for the female candidate over and ask yourself: Does it sound odd? If it does, (see summary document, appendix). As a you should probably change the terms used. result, course participants overwhelmingly gave the male postdoc a higher ranking than the female postdoc. ASKING SOMEONE TO WRITE A LETTER FOR YOU Many letter writers also need letters written on their behalf at some point in their careers. If you are asking someone to write a letter of recommendation for you, make sure you follow these guidelines: ʌ Plan way ahead. If you want good letters of recommendation, you need to plan several years in advance. It is important to have established personal relationships with potential writers early on in your career. The best letters are the ones that come from people who say “I have known this person for x years.” Try to establish strong relationships with at least half a dozen people, some of whom should be well known in your field. People who read the recommendation letter are more likely to care about the opinion of someone who has a good reputation in the field. ʌ Be specific. When you ask someone for a recommendation letter, be very specific about what you want. ʌ Prepare the writers. Once your reference has agreed to write a recommendation letter for you, a good strategy is to give the person a memo that describes the job or jobs to which you are applying and the key skills or attributes and other aspects of your background that make you well suited for the job. Other documents you should provide, depending on how well the person writing the letter knows you, are a CV, any papers published and in press, and copies of academic transcripts. You should also provide parts of the application that you have had to write yourself, such as a personal essay of your career goals or a statement of your teaching philosophy. The letter writer can speak to these items if he or she has the complete application. Making the Right Moves A Practical Guide to Scientific Management You may be asked to write a draft of the letter. As mentioned on page 1, many T I P : It is usually a good idea to ask people you heads of laboratories say this is not a good would like as references if they feel that they can idea. However, if you are asked to do it, write a “positive letter” of support for you. If they do it! say no, it is better to know it now rather than after Make sure you provide the writers with you don’t get the job. the instructions for submitting the letter and the deadline. If letters are to be mailed, give them postage-paid addressed envelopes. Also provide any forms that are supposed to be submitted with the letter. (Make sure to complete as much information as possible on those forms so that the letter writer does not have to do it.) ʌ Meet with the writers. If possible, meet with the letter writer and go through all the information you are providing. If the writer is in a different location you can talk to him or her by phone. ʌ Don’t miss the deadline. Give references plenty of time to accomplish their tasks. Ideally, approach them with the request at least three weeks before the letter is due. Once you have provided all the information, remind them of the deadline and check back periodically. Reminders will not be resented. ʌ Don’t argue. If someone says that they do not know you well enough or don’t have the time, don’t try to convince them otherwise. T I P : If you get the position, make sure that the reference writer knows it and then thank the person formally with a note. Who knows, there may come a time when you will want to ask this person for a recommendation letter again! Like many other tasks new faculty have to perform, writing letters of recommendation becomes easier with practice. But regardless of how many letters you write, each letter requires time and effort. Take the job seriously. Although a poorly written letter is not going to sink a career, a well-written one can make an application stand out from all the others.