Saturday, April 30, 2016

Dumb Study Says Unarmed Black People Less Likely To Be Shot By Police

The Washington Post’s slide into the ludicrous continues with its laughable write-up of a new study out of Washington State that boldly suggests that unarmed black people are much safer from police shootings than white people. Indeed, the paper trumpets the study with the headline: “This study found race matters in police shootings, but the results may surprise you.” You have to appreciate the now passé “knowledge gap” clickbait headline purloined from all those websites that ate the print media’s lunch five years ago. It serves as a metamarker of how out of touch the accompanying Washington Post article will be.

The crux of the article is a study — to be published in Criminology & Public Policy — that took 80 officers of the Spokane police department and placed them in a “realistic” simulator and tested their reaction times against video of actors playing white and black suspects either armed or unarmed. Over the course of the study, officers were less likely to shoot unarmed black suspects than white suspects and the reaction time of the firing officer was noticeably slower when evaluating a black suspect.
What’s the Han Solo line? “Look, good against remotes is one thing. Good against the living? That’s something else.”
To wit, empirical statistics of police shootings from 2015 via The Guardian’s database found a different story: “Police killed almost five black people per every million black residents of the U.S., compared with about 2 per million for both white and hispanic victims.” This does not take into account the “armed v. unarmed” distinction, but casts considerable doubt on drawing any conclusions from police reactions against simulators when officers feel no — real or imagined — subjective threat. Let alone gleaning anything from a study where police officers know someone is monitoring them. There’s observation bias and then there’s observation bias.
Indeed, the best line in the article notes and passively dismisses this glaring problem:
[Research Assistant Professor Lois James] considered whether the fact the officers knew they were being observed played into their actions, but she said the police did not know that race was a factor in the project.
Yeah, I’m sure a bunch of officers in SPOKANE asked to shoot at white and black people (the latter as common as unicorns in Spokane — or, more precisely, around 2 percent) would never have imagined the test could be about race. It must have seemed highly improbable that in America circa 2015-16 anyone asking you to shoot at possibly unarmed black people would ever bestudying racial bias in police shootings.
And yet here we are, with the troglodytes over at Newsbusters proclaiming scientific victory over “The Black Lives Matters folks and their enablers in the press” and the article making the trolling rounds on social media.
Most bizarre of all, the study prides itself on the fact that independent testing revealed a good deal of racial bias among the test sample... and treats this as a positive finding for their conclusion:
Perhaps stunningly, 96 percent of the nearly all white officers demonstrated implicit racial bias, with 78 percent strongly or moderately associating blacks with weapons, and zero percent associating whites with weapons. So that’s the baseline test group for the study.
No, that’s not stunning. It’s compelling evidence that the obviously artificial test was itself a variable disrupting the results and that we can’t determine how many officers would — despite their bias — do their duty and place those biases in check to make the right call as opposed to how many would let those biases run wild in the harsh light of the real world.
Perhaps basic critical reasoning skills are a lost art these days, but when research professionals gloss over basic methodology problems to back up their chosen political outcome we’ve got real problems.

No comments:

Post a Comment