Thursday, August 11, 2016

Avoiding (Unconscious?) Profiling and Microaggressions in Student Assessment


In this post, I will explore how unconscious bias and/or microaggressions can play out in the realm of the physics classroom, particularly in connection with student assessment. Below I describe two of my experiences as a faculty member where my own biases have created circumstances that are threatening or unfair (or both) to students of color.

In both of these cases, I was assessing written work from undergraduate students in the form of 3-5 pages essays outlining their understanding of a particular research question in physics or astronomy. As a scientist, I have considerable experience grading homework assignments and exams which are primarily quantitative, but more limited experience grading written work. I assigned the essays because, well, scientists write a lot, and it seems only fair to help our students develop this skill. To accommodate my inexperience, I talked with teaching and learning professionals at my institution(s) and was careful to adopt an assessment matrix for written work, which I shared with students so they knew how they would be evaluated. Despite these precautions, I still fell into a trap that I fear may snare other faculty: my own unconscious bias.

In the first scenario, I was grading 60 student essays, some of which were well written, many of which were not. As I worked, one particular essay set off alarm bells — it didn't “sound right”, i.e., it struck me as not sounding like it was written by the student whose name appeared on the assignment. It was probably not a coincidence that the student was a young black man on my institution’s football team. I spent some time trying to ascertain whether some or all of the essay was plagiarized, but did not find any particularly strong evidence. I discussed this with colleagues who noted that faculty’s “instincts” are often correct in these cases and I should pursue it if I felt strongly.

So I asked the student to send me evidence that the work was his, e.g., an earlier draft, an outline, whatever he had that would help me see his train of thought or the leg work that went into creating the essay. He put me off for several days because he was on travel with his team, then sent me a document with lots of comments embedded, from his coach. I didn’t love this (the comments from the coach, that is), but since I encourage my students to have someone else read and comment on their work, it wasn't totally out of line. But then I looked at the electronic authorship on the Word document and it wasn't my student’s name. Now my hackles were up and I asked my student to come talk to me. We discussed the series of events and the strange name on his document. It turned out that his laptop was part of a computer sharing program supported by my institution, designed to insure that all students have access to the resources they need for their school work. That other, electronic name belonged to the previous owner of the computer. At this point my student was freaked out that he was going to be censored for misconduct, and I was painfully aware that I had singled out a student of color on the flimsy evidence of my “intuition”. I thanked him for his answers to my questions and let him know there would be no further inquiry.

This was an all around unpleasant situation, one of my own making, and one with considerably larger negative impacts for my student than for me: I had questioned his integrity in what may have been one of the only science classes he selected during his undergraduate education. Me dropping the inquiry was small consolation for that stress. And even if the plagiarism had been clear cut, my scrutiny was not equally applied to all of my students. It had, in fact, been handed out to only one student, a young black man. This was a textbook example of an academic microaggression, precipitated by me, a person who tries hard to create supportive learning environments.

Fast forward several terms and I am reading another set of student essays, now for a different class. Once again, a particular essay bothers me, for reasons I can’t fully articulate. This time I am reading the essays without looking at the names, but when I do look at the author, it is a person of color. This time, I am determined not to repeat my previous mistake. So I submit ALL of the essays to a plagiarism checker. For anyone who has tried this, you know what a time-consuming pain it is. The checker picks up a zillion little things that are clearly not a problem, everything from turns of phrase to titles of books in the reference section. I had to look each essay over one at a time to remove all the spurious detections. After this review the student’s essay that had tickled my professorial intuition revealed only a tiny amount of text that overlapped archives. A completely different essay however, one which I had already marked with a 95% (!), written by a white male student, had more than 20% overlap with Wikipedia. You read that right, Wikipedia. Seriously!?

The moral of this story, if there is one, is that we must hold ourselves accountable. Our students rely on us to be fair and evenhanded. If our own unconscious biases are guiding our “instincts” we must learn not to trust our instincts or to re-train them. It is not acceptable for us, for me, to hold different students to different standards.

Of course, plagiarism is a problem, and I am all ears for better strategies for addressing it. My current best idea for essays is to institute more peer review. I already have my students reverse outline each others’ work and going forward I will have each student run their peers’ work through one or more plagiarism checkers, but this still doesn’t get at plagiarism of homework solutions or in other arenas. If you have solutions to share, or would like to contribute to a discussion about how we can strive for more equity in our physics classrooms, please contact me.

So there you go, I've provided clear evidence that my intuition cannot be trusted. How confident are you in yours?


A few basic definitions:

* Unconscious bias: attitudes and stereotypes that affect our understanding, actions, and decisions in an unconscious way (without awareness), e.g., see a previous CSWA blog post by Meg Urry.

* Microagressions: the everyday verbal, nonverbal, and environmental slights, snubs, or insults which communicate hostile, derogatory, or negative messages to target persons based solely upon their marginalized group membership; these can be either conscious or unconscious. Jessica Kirkpatrick has previously posted on this blog about Microagressions.

1 comment:

  1. Some of these WIA posts expose an utter lack of understanding of the lives of American students today is US higher education.

    "It turned out that his laptop was part of a computer sharing program supported by my institution, designed to insure that all students have access to the resources they need for their school work. That other, electronic name belonged to the previous owner of the computer"

    Though this (ostensibly) happened at a Canadian university(?), the sentiment below still applies:

    Did you know that around 72% of Americans live paycheck to paycheck? And that in 2015, over 70% of US college students had to have part-time jobs to attend university? That many (lower-income) US students are able to go to university via scholarships given on the basis of athletic prowess?

    Yet, the assumption here seems to be that all students naturally who have a laptop with a licensed version of Microsoft Word. Based on my reading of this case, (and I don't mean this to be ad hominem or a personal attack towards Prof. Haggard---I notice this in many WIA posts) in terms of "unconscious bias", the glaring problem seems to be that Prof. Haggard is an upper-middle class American who grew up in the 1990s, and has no idea how fucked up the lives of her students in 2016 are.

    Naturally, there's a discussion to be had here about race. But my reaction upon reading this article is "You were picking up the poor kid! That's your bias!"

    ReplyDelete