Nov 30 2011 Published by under letter of reference

Questions abound about what should and should not go into Letters of Reference. I will not include any particular reader question here, but will try to hit the major points that commonly arise.

Of course the purpose of the letter and the nature/length of the letter will vary depending on the purpose and the personalities involved, but there is a certain sameness to these things as well, whether the letter is for an undergraduate applying for a summer internship or a postdoc applying for a faculty position.

Over the years, I have marveled at some of the weird things that people put in reference letters. I think the weirdest items appear in letters for undergraduates applying for internships or graduate programs because the letter-writers:

(1) may not know the applicant very well and struggle (in some cases, mightily and inappropriately) to find something to say other than "Jane got an A in my class"; and

(2) may know the applicant very well indeed and may somehow lose perspective on whether potential research advisors want to know that Jane was a great babysitter for the letter writer's 7 children (I personally do not want to know this. There are people I would trust with kids but not research, and vice versa.)

Just a few examples from the FSP and SP archives of strange and possibly inappropriate things I have seen in letters of reference for academic positions, awards, tenure, and promotion over the years:

- phrases like "Applicant X is one of the best female graduates of our department";

- or this: "He and his wife have spent many vacations in Bavaria, organizing their walking routes to coincide with the locations of breweries."

- "In my opinion, Applicant X is an excellent scientist. Now let me tell you about my credentials. Attached is my CV."

- "Applicant Z's Christian faith has helped guide her through a rigorous academic program."

- "Molly is such a responsible and mature person that my wife and I have repeatedly trusted her to care for our 5-year old and 2.5-year old when we have a 'date night' or a social function that is not suitable for young children."

- "Dr. X has come from a distinguished academic line. One of his committee member's advisor's advisor's advisor was awarded the Nobel Prize in 19xx."

OK, enough of that (for now). Writing letters is difficult, no matter how little/well you know the applicant. No one is perfect, so a common question by letter writers is how/whether to describe or hint at some of these imperfections.

Arguments in favor of writing a positive letter that has a few minor mentions of reasons why the applicant is mortal, even if you mostly think this person is awesome:

Letters should convey useful and accurate information. If someone's imperfections are relevant to the position for which they are applying, wouldn't the letter readers want to know this? Your credibility is at stake as well, and therefore your ability to advocate for others in the future.

Argument against mentioning these unless they really are major, fatal flows (in which case there is the issue of whether you should have told the person who asked you to write the letter for them that you couldn't write them a good letter):

Many (most?) letter-writers don't write anything negative in letters, so if a particular letter-writer does say something negative, however mild, that may doom the applicant's chances because all other candidates are apparently perfect (even if committees/individuals reading the letters know that the letters are likely to be somewhat incomplete in this way).

I think you've got to do what you think is right in each case, and just be as straightforward and unambiguous as possible. I have spent way too much time in committee meetings listening to people try to divine what is meant by a possibly somewhat ambiguous turn of phrase or choice of words -- is this a Red Flag intended to signal that the candidate is fatally flawed as a human being and a scientist, even though the rest of the letter is entirely positive, because the letter-writer didn't want to commit on paper to writing a major criticism? Or is that phrase just what it seems; a simple statement of something reasonably positive?

But maybe people will try to 'read between the lines' no matter how unambiguous you think you are being in your writing, and therefore it isn't worth anyone's time to try to psych the situation out.

These are some general "rules" that I try to follow for myself when writing letters (please add to the list with your own personal LoR-writing rules):

- I write what I think is fair, relevant, and useful in the context of the letter. I tailor each letter to each applicant and letter destination/purpose. I back up opinions with examples or other information.

- If the letter request contains specific questions or topics that should be addressed, I try to answer/address these as much as possible, unless I think the question/request is unreasonable. I have written before about requests to compare someone with their peers; that can be an extremely challenging request, fraught with potential for unfairness.

- If I don't have much to say about someone [and they are aware of this fact, but don't have other/good options for letter-writers], I keep the letter short. I explain that I had limited interaction with the candidate, so my short letter will be understood in that context, rather than that I was too busy to take the time to write a decent letter. I think some of the stranger letters I have read arose when the letter-writer started fishing around for things to say to bulk up a letter.

- I avoid personal information (hobbies, babies..) and irrelevant information about personality. Studies have shown that unconscious bias creeps into the adjectives we choose to describe the personality of female vs. male candidates, so I fight the urge to describe someone as "nice". If the person in question gets along well with others, there are other ways to explain that, such as with examples of successful collaborative work. I also don't think it is relevant to mention whether someone has a sense of humor; I think this is more common to describe in letters about male candidates than about female candidates, even though I doubt if the men are actually funnier than the women. [Yes, I know there are studies and debate about this.]

My letters are by no means perfectly crafted and compelling vessels of information, but I think it's important to try to write a good, useful, convincing letter. This takes time, of course, but it is time well spent.

What are some other Rules to Live By when writing a Letter of Reference?






14 responses so far

  • sciencecanary says:

    One of the hardest things to write about is writing. Someone may be a great researcher, person, speaker, teacher, and all that, but struggle to write. And yet they want to have an academic career in which they have to write papers and proposals. If I had to drag every single paragraph out of them, painfully for both of us, and edit their thesis chapter(papers) 17 times until there was more of my writing than theirs, that is relevant information for a search committee for many academic jobs. Sometimes I ask my advisees "What do *you* think I should say in this letter about your writing abilities?", but this hasn't been very useful because no matter how much someone struggles, they always seem to think that this problem will magically go away once they have a job, and therefore I shouldn't mention it. So maybe the problem is me? Maybe I have such strict writing standards, that, once freed from this pressure, their prose will flow freely and copiously? Maybe, but in the meantime, what do I write about their writing?

    Perhaps this is where people respond that there are many more PhDs applying for academic jobs than there are academic jobs, so the weak should be culled and only those who can and will write should survive. The response to that of course is the point about how not all letter writers mention such flaws, so why should particular people be at a disadvantage? And also in my field now there are quite a few academic jobs, so the PhD-oversupply of other fields is not relevant.

  • Polytrope says:

    My impression is that people calibrate for nationality of the writer as well. Americans are always unremittingly positive (which as a European reading those letters I find a complete nightmare since it leaves you almost no useful way of comparing). I usually mention points that need development and which a future employer might find useful to be aware of. So far this hasn't stopped anyone I've written for from getting a job, so hopefully they are calibrating for 'realistic European'!

  • mathgirl says:

    Thank you FSP for the timely post! I have an idea for another post: how about reading recommendation letters? These days I have to read tons of recommendation letters and it's hard to interpret them. Like it's mentioned above, some people write generic positive letters, others make sure they always mention a flaw (if small). Also, what to do when the same person writes letters for two candidates and makes it clear that one of the candidates is better than the other?

  • BugDoc says:

    My rule of thumb is to be positive about everything that I can honestly be positive about, but clearly flag issues I think will affect performance. I had a student who was amazing at the bench and had a very strong work ethic, but was a terrible writer. He tried very hard to improve his writing and we worked together extensively on it, but he really struggled. I couched this problem as "X is working hard to improve his writing skills and shows great motivation to excel in this area, as he does in all other aspects of his research." I hope this communicated the problem without being unduly negative. Like mathgirl, I'm also interested in hearing how different people interpret letters.

  • RespiSci says:

    I have sat on MSc, PhD and post-doc fellowship selection committees and have reviewed many letters of reference. In this committee we didn't care when the referee would write a paragraph about how excellent the student was academically as we had access to the full academic transcripts and the referees should have known this. Most of the candidates were stellar academically, and had solid research proposals (and publications) so in order to separate the wheat from the chaff we needed to know more about the candidate as a mortal. We wanted examples of how they may have showed passion for research (besides working hard and being dedicated), examples of teamwork and being independent and yes, we wanted examples of personal aspects of the candidate. Often in the reference letter we would learn of some personal hardships faced by the candidate that they overcame, or how the candidate was doing volunteer work, organizing events and other non-science activities. In today's research age, the ability to collaborate, multi-task and go above and beyond is demanded of in academic researchers. So we had to select those candidates who could do it all (because we only had so much money to spread around).

  • Colleen says:

    Show, don't tell! Examples trump descriptions.

    I took the controversial route of seeing my grad school recommendation from my advisor before it was sent out, something I had to declare on my applications and which came up in a few interviews. The reason being that my advisor didn't want to make a mistake in my letter. It had the added benefit of me being able to request that he include more specific examples of my awesomeness instead of more general descriptions, i.e.:

    Version 1: "Colleen is responsible for the day to day management of a major research project."

    Version 2: "Colleen manages our NIH grant, "Title and Number." She is responsible for all recruitment, screening, scheduling, data collection, data management, and a substantial portion of data analysis. She has recruited X people, collected data from X people, presented her work at ABC and XYZ conference...etc."

    BIG difference. Can you guess which style of letter got me waitlisted but no acceptance, and which style got me 6/15 Ph.D. acceptances?

  • Am says:

    "Applicant X is one of the best female graduates of our department"

    Let's assume that the top graduates of a certain department are mostly male and a given female graduate is not between the top graduates of the department, although she is (one of) the best female graduate(s). Would you find it inappropriate to mention that point in the reference letter? Do you think it is better for the letter-writer to write that the applicant is an average/mediocre graduate instead? Just curious to know.

    Needless to say, these are just hypothetical assumptions and it has never been the case that the top graduates of a certain department are mostly male.

    • Dr Moose says:

      Even if you mean well and are trying to find a way to say something positive about these hypothetical women, would you say something similar about grads from, say, India or China? Imagine seeing in a letter "Applicant X is one of the best graduates from India of our department." That would be inappropriate. So why would it be OK to say that for women?

  • jen says:

    As a rule, I usually don't usually say anything outright negative, but I definitely calibrate my adjectives.

    X was a scientist in my laboratory.
    X is a good scientist.
    X is a superb scientist.
    I have trained 15 postdoctoral fellows who have gone onto independent positions, and among this group X stands out as the most excellent scientist I have ever worked with.

    I also look for this when I am reading letters.

  • acat says:

    It's funny how different US and European letters tend to be. I agree that US letters are filled with all sorts of hyperbole and amazing adjectives, but I think the terse European letters (typically) go too far the other way. There are many exceptions, of course, to these general characterizations of letters from both continents, but a brief European letter that basically just says "Yes, I know Applicant and I think this person is actually quite good (and here are my credentials)" is not very useful.

  • A Physicist says:

    I try to highlight specific abilities:

    X performs intricate calculations well;
    Y is very original in his approach;
    Z is deeply knowledgeable about the work in this field;

    As for writing abilities, I have said things like

    X works best in small groups or with a coworker, who can put in words what X has found through calculations.

  • Joel says:

    Maybe the typical terse European letter looks like this:

    X is a student who is off to a fairly decent start in his/her career. The NSF awarded us $100K for our project and X came up with ideas that enabled us to spend $200K. We are a top notch lab so while the truly star candidates would be expected to produce at least a 400% deficit, but X does clear the minimal standards with a 100% deficit. By the way, would your lab care to purchase some 10 year bonds from our lab at a friendly 1% rate to finance our deficit? I remind you that if our famous lab went belly up, it would have terrible consequences for worldwide science.

  • [...] checkout FSP’s recent post in Scientopia about whether or not to say anything negative at all, other “Rules,” and a helpful [...]