| title | author | date |
|---|---|---|
EDI brain dump |
Ben Bolker |
8 December 2022 |
\includegraphics{cc.png}
Another brain dump of thoughts after attending a diversity/equity training workshop at McMaster.
I understand that the university needs to make sure that people stay up to date with training, but it feels terribly wasteful to spend hundreds of person-hours on this when most of it is not new (the facilitator in my session apologized at least four times for the repetition). Would it be possible to replace the repeated training with something like materials to review plus a quiz, then ask faculty to submit questions and engage in a Q&A session?
Note that the training videos from Facebook don't have an option for closed captioning. This seems particularly ironic since the university's biggest gap category is disabled persons ...
I know that socioeconomic status is not in the list of protected classes and as such will be lower on the list of cases to consider, but it would be interesting to mention or discuss - seems important in many of the same categories (diversity, equity, barriers to inclusion) as the groups we discussed. Where it intersects with other categories (indigeneity, racialized identity) it may get addressed, but otherwise it may not be considered by hiring managers/committees.
Thank you for censoring the cases with fewer than 6 individuals in the gap analysis data (e.g. indigenous assistant professors at McMaster). I don't know if this was done for privacy reasons or because such small samples would be too noisy to be useful, but it's a good idea in either case.
I will be very interested to see an analysis of the longitudinal data for representation of women at different ranks in the tenure-track stream over time (i.e., whether the underrepresentation of women at the full professor level is primarily a cohort problem or represents a leaky pipeline/failure to progress). My very anecdotal/personal experience would suggest there's not a pipeline problem for female professors within McMaster, but I recognize how unreliable and limited those impressions are.
I've said this before, but if the goal is really to achieve bias-free screening in the early stages of the hiring process, is there a way we can do blinded reviews at the first stage? There are plenty of challenges to this (we would have to eliminate gendered pronouns and references to the candidates' names in all components of the applications, including letters of reference), but it seems at least worth considering whether they could be overcome.
(i.e., allowing one positive/preferred or negative/disliked aspect of a candidate to affect one's overall evaluation). These are cool terms I was unfamiliar with. In my experience it comes into play mostly at the level of individual sub-field preferences in broad searches. When someone really wants a particular candidate because of their field of study, they will strategically emphasize the criteria on which that candidate stands out (e.g. teaching experience, publication record, gender, etc. ...) and de-emphasize the other criteria.
Contentious, lots of research. Please recommend its use for self-assessment with great caution (self-reflection yes, self-assessment no): psychologists are still arguing hotly whether IATs are a valid measure of individual attitudes, how strongly they are correlated with explicit statements of attitudes, how strongly they are correlated with behaviour toward people from different groups, etc..