Gender Bias, Marking, and Hand-Written Exams.

I have barely done any marking yet, but this struck me as I was distracting myself with observing my colleague marking hand-written exams: I (bored) picked up an exam, read a few notes, and made the following comment: “listen to this, her grammar is terrible:……”. Then it struck me – why did I assume the author was a girl?????????? Answer: pretty, neat, round handwriting…….

So, here’s my worry: we know how implicit bias acts in the context of judging CV’s, reviewing papers, judging comments, etc; when identical papers/CV’s/etc are labelled as originating from either women or men, they are generally judged to be of comparatively lower quality when marked as originating from women. That is why it is so important that these things are (ideally) done anonymously. Now my experience with this exam paper suggests that, despite not knowing the author’s name, I had made a very quick, unconscious judgment about the gender of the author based on her handwriting. This judgment may not always be accurate, but it if it is accurate often enough (and a quick flick through the pile of papers resulted in many papers that me and my colleagues unhesitantly judged to be either male or female), then this is a very pernicious way in which exam-marking can be distorted by implicit bias…

And most exams I have seen are handwritten, which suddenly makes this quite a worry? Thoughts?

14 thoughts on “Gender Bias, Marking, and Hand-Written Exams.

  1. Nice post – I think you’re absolutely correct about this, wahine1.

    I had an interesting conversation with my daughter over the holidays about more or less the same point. She had just discovered an old ornament with my name on it, in handwriting from some n decades ago, when I was, oh, about 13. The letters were round, neat, cute. A big hollow bubble for the “i”. Purple ink. My current handwriting is an almost-illegible thin scrawl, and my daughter wanted to know when (and why) my handwriting changed.

    The ‘when’ was easy: sometime late in secondary school, one or two of the fellow students in my all-male group of peers on the math team teased me about my “girly” handwriting. Real mathematicians don’t write in purple, they argued.

    The ‘why’ question was a little harder. The teasing session was on the way back from a victorious meet in which I had answered some key questions. I could’ve simply pointed out the obvious: that I was real, that I had more of a claim to being a mathematician than they, and that I wrote in purple ink. But instead I ditched the purple pen and, with a few hours of practice over 2 or 3 days, deliberately transformed my handwriting to a thin, slanted scratchy print, very much like the writing of one of the well-respected male math teachers.

    Why was this my response? We all choose our battles, and it might have been simply that this was the path of least resistance. But, in retrospect, I wonder whether it wasn’t also because I had become sensitized to the insidious effects of implicit bias, and realized in some pre-verbal sense that it would be prudent to de-feminize my writing — just as I had already de-feminized other aspects of my presentation of self.

    Which leaves me wondering: what are the factors that influence these sorts of prima facie unreflective reactions to the presence (or perceived presence) of implicit bias? Why do some girls change their handwriting, while others don’t? Are some girls/women more sensitive to implicit bias and stereotype threat? Are there some contexts in which implicit bias and stereotype threat are more salient? What sorts of factors might mitigate the effects, prompt and encourage reflection, redefine the path of least resistance?

    I don’t think I can easily change my handwriting again at this point in my life, but I am most definitely going to purchase a purple dry erase pen this weekend to use in lectures next week. Pink, too, if I can find a nice vibrant shade.

  2. This certainly looks like a serious problem. I’m not sure whether my students get the handwriting instruction that makes pretty handwriting available, but on thinking about this, I realized that I seldom give in-class exams any more.

    It’s also too possible that some students try to write prettily because they think it will incline profs to give them a higher grade.

  3. Not only is it a worry, but I think the worry moves down to a much more detailed level than simply gender. I’d venture to say that there is a worry of correlated particular personality types within gender (bubbly handwriting, or macho handwriting, or sloppy/slacker handwriting, or studious/smart handwriting) to handwriting styles.

  4. I started anonymizing my exams a few years ago and immediately realized, when I went to grade the first batch, that the handwriting issue would be a problem. Interestingly at first the only times I found myself making a quick, automatic judgment about the gender of the student, I took it to be a woman. But then I noticed *that*, and every time I picked up a new exam I was incapable of resisting a guess about the gender of the student. The whole experience was very strange — my awareness of the gender of the writer felt heightened way beyond what it would have been had I simply known his or her actual identity.

    Obviously this problem doesn’t emerge when written work is typed, but that’s often just not practical where I teach. I still think anonymizing is worth it but I’m also aware that it doesn’t correct for all the implicit biases.

  5. I remember, way back in the Dark Ages when I was in high school, a fellow student who wrote all her reports and exams etc in a lovely multicolored rounded fancy “girly” handwriting, complete with heart dots on the “i” and flowers drawn in the margins. She always got an “A”. Content was nonexistent, and evidently not important… I hated her.

  6. I have all students type their exams. Take-home format, with their name printed on a blank, LAST, page of the exam they hand in. This negates the handwriting issue (as well as the related concern that you may recognize a student’s handwriting).

  7. If a blind teaching assistant/instructor was advised that marking her exams should be “(ideally) done blindly,” do you think this would make her susceptible to stereotype threat? That is, would it affect her chances of doing her job well?

  8. For exams last semester, most of my students printed their words. A few used handwriting, but from what I could tell from looking at their names afterward, there didn’t seem to be a division along gender lines between cursive and print. There did seem to be a geographical division, however–I had a few students from India, and they all used cursive.

    When I write things out by hand, I too use cursive (I’m male). I suspect that has more to do with our elementary school education/highschool writing assignments than anything, however.

    As far as neatness goes… I don’t know if there’s a correlation. I know that when I’m under time constraints, both my printing and my cursive suffer (I tend to write a lot, so to have enough time to write it all I sacrifice legibility). Without any real time constraints, it’s neat. Perhaps the same applies to others more broadly–I don’t know.

  9. My daughter took Michael Sandel’s “Justice” course last fall (justiceharvard.org), and all the papers were submitted with no name – just a student ID – to turnitin.com. They were somehow parsed out to the TAs for each section, and the TAs graded the papers with no way of knowing whose it was. There were also TA leaders of some sort who were not involved in the actual grading but were available to discuss and review drafts with individual students.

    This doesn’t necessarily eliminate all implicit bias, because there may be certain modes of presentation/thought that are gender-based. It also doesn’t eliminate bias that may enter into the feedback that students receive to ideas presented in the section discussions. But it does eliminate any implicit bias based on handwriting and/or explicit awareness of a student’s gender.

  10. It may be that having neat handwriting is an advantage. It is pretty irritating reading an angular scrawl.

  11. I agree with Tina. If there is really some bias at work here, I suspect that it favors those with neat handwriting.

  12. Thanks, Shelley. Post now changed.

    (Wahine – hope that’s ok – I’m not sure whereabouts in the world you are, so as I’m awake and online now, I altered the post.)

  13. I agree, this is a huge problem. It is time for change, embrace new possibilities & balance to all this 2012…….. Embrace your gender feel free to be who you truly are & shine

Comments are closed.