Would it be appropriate or inappropriate for me to apply for positions at HBCUs in the US, seeing as I’m a white guy? A couple will be hiring for the exact subfield I study, and while I want to apply for every position in that exact subfield (of which there are only finitely many) at the same time I obviously don’t want to barge into an institution that exists to provide support and mentorship and networking opportunities for a marginalized community of which I’m obviously not a member. What are the norms for this?
the most extraordinary sentence i’ve ever read in a quantitative research paper?
Nosek BA, Banaji MR, Greenwald AG. Harvesting implicit group attitudes and beliefs from a demonstration website. Group Dynamics. 2002;6:101-115.
OH MY GOODNESS this paper is shady for a whole lot of reasons (check out the names in Appendix A to see just how strongly researchers are trying to adjust things to get exactly the responses they want to see) but the absolute worst aspect is that they don’t bother to test for statistical significance.
I mean, okay, under the 19th-century Platonic ideal of statistics where you have a finite number of observations and they all measure some fundamental underlying truth plus or minus a random error where successive observations are perfectly uncorrelated but have identical error term distributions, then it’s perfectly reasonable to be like “oops my error bars are too small to matter”. But unless you’re doing, like, particle physics or something, this isn’t the way the world works. There’s certainly going to be massive correlations between observations by the same person and probably also observations between respondents who chose the same task.
And hey guess what we’ve had ways to deal with that extra correlation for several decades now! And basically every statistical software in existence can handle this perfectly with no problem! And so if you’re publishing a paper that pretends that all error terms correspond to some ideal that seemed like a reasonable approximation in the Victorian era when professors would hire a roomful of young ladies to crank out statistical results by hand, then one of the following two things has happened:
- You’ve somehow made it through decades of employment in quantitative social sciences without picking up any of the intuitive out-of-the-box statistical techniques for correcting for group-level correlation.
- When treated more appropriately, your results are incredibly shady and the error bars are gigantic, and you’re counting on your readership not to spot your shenanigans.
And yeah wow neither of these is reassuring whatsoever.
If you’re blue and you don’t know where to go, why don’t you go where fashion sits?
Like, the writ is dropped and the Parti Québécois are heavily favoured to win the election, as they’ve successfully turned it into a referendum on protections for the white francophone majority at the expense of everyone else. The two biggest policy priorities of the PQ a ban on hijab, turbans, kippot, and kosher and halal food at all schools and daycares and hospitals and government offices but a giant crucifix in the legislature and on the top of Mount Royal, plus a requirement that every business with at least eleven employees conduct all conversations between management and employees in French. The government dresses these policies up as protecting the fragile livelihood of the hard-done-by white francophone majority (who only control all of politics and business and media in the province) at the expense of, I dunno, the 5% of the population whose culture differs substantially from that minority.
Right, anyway. There are maybe four ways the election could go right now. All these probabilities are pretty much guesses, at this point:
- Liberal minority (5%). The party overwhelmingly supported by non-francophones comes back to power. Expecting dithering policies followed by incompetent public responses, but, you know, no overtly racist flagship laws. Only way this happens is if the CAQ (a right-wing Quebec nationalist party) collapses and voters go back to the Liberals.
- PQ minority (20%). Basically the status quo. Expect very little to happen and another election as soon as the party feels a little better about its chances.
- Small PQ majority (60%). By far the most likely outcome. Expect all that awful promised legislation to get passed, big dramatic showdowns with half the big hospitals and universities and civil society-type groups in Quebec, and a big showdown with the federal government and the Supreme Court culminating in Quebec invoking the notwithstanding clause. So, you know, ~just Quebec things~.
- Big PQ majority (15%). Like, at least ninety out of the 125 seats. This wouldn’t actually require a landslide in the polls, because Liberal votes get wasted by their geographic concentration in a few ridings. Expect serious talk of a new referendum if this happens, even though the PQ has backed off on secession talk for the beginning of their campaign.
Anyway, the election is on April 7 so only about five weeks to go for all of this.