Recommendation Letters for Women Can Be Chutes
The author, Jessica Stark, referred to the effects of letters for women as "chutes," in the Chutes and Ladders sense of the word. Not good. Language describing women applicants tends to be emotic while that describing males tends to be agentic. That is, men are described as do-ers and women are described as feel-ers as shown in the hundreds of letters they analyzed. This difference is not just in letters by men. It also occurs in letters by women.
After first noticed this problem, based on a different study some years ago, I looked at letters I was writing for students and colleagues and found I usually included a mix for both genders, but the emphasis was in this same wrong direction: more emotic for women, more agentic for men. So I tried to scrub this out of my recommendations and I hope I've done it fairly well.
Has anyone had experiences with reading recommendation letters recently, and if so can you comment on this effect? I had hoped the trend was away from such language.
What do you think?
A Today, it's rare for women to be described as emotic and not agentic.
B In some letters it's this way, but some letters do describe women as agentic, and some are mixed.
C Totally, this is how letters work. Women don't get described as future Nobel laureates. Men don't appear as nice cooperative colleagues. Stereotypes unconsciously direct the word choices people make re female vs male applicants.
Hi Phoebe,
Interesting questions, ones that probably require study to see if there are trends in what actually happens as opposed to what people think they recall. Like Helen, I found that the places that I applied suggested certain qualities they wanted in the letters, e.g. someone who can speak to outstanding research potential, someone who knows your teaching, etc. Female vs male was not part of the suggestion, and at the time, almost everyone of note was male so I had no female referees. I hope things are better now. Certainly women I know seem to write plenty of letters!
cheers,
Laura
Phoebe: I was told to get letters from the most powerful and well regarded scientists I could. I ended up with all men. I don't think its meant to deliberately exclude women. And of course in certain instances I had friends who had 2/3 letters from famous women as part of their job package. This didn't seem to hinder them. So I think it boils to impact and influence of the writer. And there are fewer of those types of individuals who are also women. Get more powerful women and you'll see them being asked for more letters...chicken and egg problem again.
AWIS has an NSF grant to improve recognition of women in disciplinary societies, and one issue we are working on is how to improve the impact of letters of recommendation. The conventional wisdom is that it is better to get letters written by men rather than women - but are there any data on this? Most of the limited evidence is that both women and men write letters tending towards use of emotic descriptors for women and agentic descriptors for men. Does all of this mean there is no advantage for a women to get letters from women, and a possible disadvantage?
Here is one of the references about recommendation letters for women vs men that I had seen earlier:
Trix and Psenka, Discourse & Soc 14:191 2003
• 312 letters of recommendation for medical faculty
who were successfully hired.
Letters for women:
O Shorter
O More letters for women with “minimal assurance”
O More gendered terms in letters for women
O More letters for women included “doubt raisers”
O Men more frequently referred to as “researchers”
and “colleagues”. Women more frequently referred
to as “teachers” and “students”
o Women –4X more references to personal lives
o Women -Fewer standout adjectives(“outstanding”
“excellent”) and more grindstone adjectives.
Cheers,
Laura
Hi Helen,
That cracked me up. Had to be sure he changed the he to she, huh? Could have made people think you'd had a gender change, since that's getting more frequent these days. But nice that he used all the same adjectives. However, it would have been better had he been specific to each of you, of course.
cheers,
Laura
Laura: There are some good references on how to write a good reference letter:) I pull them out to review before I write any letter, but I wonder if with all the pressure to get everything else done if most people have the time? It may be easier to copy/paste. My postdoc advisor was infamous for that:) You had to proof the letter to make sure he changed the he to she. At least we all got the same adjectives regardless of gender...
Hi Helen,
I've seen this too; when I call a referee I get quite a different picture of the person than his/her letter gave me. I wonder if the letters are kind of stylized and knee jerk, and also if they contain selected protective language to avoid any potential lawsuits. I remember some quite bland letters for Biochem/Mo Bio NSF Graduate Fellowship candidates when I was on the selection panel. I wanted to call, but of course in that context it cannot be done.
cheers,
Laura
B. The letters are often skewed but then you call to check a reference and get a different story on the phone. I've experienced men being described as more emotic and women as more agentic in conversation rather than letters. I'm not sure why that is? Perhaps it's the dynamic conversation aspect? I will say that I have NEVER seen a letter for a man where emotic terms were used. So there is something going on here...
Yes, postdoc cat, I do think women who are described as aggressive, dominant, and outspoken would sound scary to a search committee. YES, it's unfair, but that's what we have to deal with.
cheers,
Laura
But don't you think women described by the words on the agentic list would sound like they were hard to get along with? Would people still want to hire them? Some of those words work for me, like confident and intellectual. But others sound kind of repellent.
Here are some of the actual terms that the study found used in each category, FYI, in case you get to see your recommendation letters, so you can check the types of terms used to describe you: "Words in the communal category included adjectives such as affectionate, helpful, kind, sypathetic, nurturing, tactful, and agreeable, and behaviors such as helping others, taking direction well, and maintaining relationships. Agentic adjectives included words such as confident, aggressive, dominant, forceful, independent, daring, outspoken, and intellectual, and behaviors such as speaking assertively, influencing others, and initiating tasks."
cheers,
Laura
I don't get to see the letters people send out for me, but I see some for people applying to be postdocs and I'd say C is right on target.
I am not sure I agree. Maybe give him the article and tell him it made you worry about his letter for you. Even if he dismisses it, I bet he'd think twice about his word choices next time.
I'd consider giving him a copy of the article on it with the data in it and say you're surprised this is still true? I wouldn't start by critiquing his letter for you.
Any advice about how to talk with my recommender about this problem? He's pretty sympathetic overall, but this could get awkward fast.
@postdoc cat, yes, that's a problem.
This is another version of the problem that women are typically asked to be on committees about PEOPLE or soft issues, while men get the committees about MONEY.
And students are much more likely to complain about a woman professor who's tough, than a man, since they "expect" a woman to be fuzzier.
So my answer is somewhere between B and C
One of my advisors gave me a copy of the letter he had been sending out, and it's a mixture but with emphasis on me as a lab citizen. But I have more first author papers than anyone else in the lab and I think I've contributed ideas, not just cooperation. I thought it was a strong letter, but now I see it could cause problems.
