Comment by revelio | Hacker News Reader

Comment by revelio | Hacker News Reader

Parent Full thread

That retraction was itself political in nature:

https://www.wsj.com/articles/i-cited-their-study-so-they-dis...

My June 3 Journal op-ed quoted the PNAS article’s conclusion verbatim. It set off a firestorm at Michigan State. The university’s Graduate Employees Union pressured the MSU press office to apologize for the “harm it caused” by mentioning my article in a newsletter. The union targeted physicist Steve Hsu, who had approved funding for Mr. Cesario’s research. MSU sacked Mr. Hsu from his administrative position. PNAS editorialized that Messrs. Cesario and Johnson had “poorly framed” their article—the one that got through the journal’s three levels of editorial and peer review.

Read the correction they issued:

https://www.pnas.org/doi/10.1073/pnas.2004734117

"As we estimated Pr(race|shot, X), this sentence should read: ‘As the proportion of White officers in a fatal officer-involved shooting increased, a person fatally shot was not more likely to be of a racial minority.’ This is consistent with our framing of the results in the abstract and main text."

In other words, the thing being fought over was the wording of a single sentence that didn't change their actual conclusions, which are consistent with the other evidence. But then later:

"Despite this correction, our work has continued to be cited as providing support for the idea that there are no racial biases in fatal shootings, or policing in general."

Yes, because their correction in their own words "is consistent with our framing of the results".

This whole incident is just proof of what I'm saying - the American left systematically attacks anyone who tries to point out the actual data here, which is why people don't hear about it. This particular paper led to a Twitter mob and someone got fired, but it wasn't due to bad methods, it was because student activists were upset about the "harm" it caused:

https://www.wsj.com/articles/a-twitter-mob-takes-down-an-adm...

>> you're way more likely to be pulled over while driving if you are black than if you are white.

Aren't you now doing the things you're claiming these Michigan researchers did? Where's your control for actual crime rates?

> the American left systematically attacks anyone who tries to point out the actual data here

And you're systematically ignoring literally everything else I've said, including an article pointing out the multitude of errors included in the study, references to other unexplained disparities, and I might also just add:

> Cesario and Johnson write that their "decision had nothing to do with political considerations, ‘mob’ pressure, threats to the authors, or distaste for the political views of people citing the work approvingly."

Even if police shootings aren't evidence of racial disparities in policing, there are a _lot_ of other disparities.. which are certainly worth considering if you actually care about the data.

But if your evidence consists entirely of vacuously parroting Oped pieces by a right wing political commentator who "claimed that 95% of outstanding homicide warrants in Los Angeles were for undocumented immigrants and that 75% of L.A.'s most wanted list comprised undocumented immigrants", then.. I don't know what to tell you.

You believe the retraction was done for legitimate scientific reasons instead of for political ones just because the people who did the retraction said so?

>Yes, and that retraction was itself political in nature:

>https://www.wsj.com/articles/i-cited-their-study-so-they-dis...

Indeed. The chronology is important:

1. Cesario et al [1] received some criticism [2], but they stood by their findings [3]

2. A student union / twitter mob was not happy with the findings of the study [4]

3. The resulting political pressure led to the ousting of Stephen Hsu from MSU, who had approved the funding for the study [5]

4. Only then retracted Cesario et al their study, while still standing by their findings [6]

______

There is a similar study done by economist Roland Fryer [7], while at Harvard.

Summarized, the two studies find the following:

Cesario et al:

Per interaction with the police, civilians have roughly the same risk of being killed by police gunfire, regardless of their ethnicity.

Fryer:

- Per interaction with the police, civilians have roughly the same risk of being the target of police gunfire, regardless of their ethnicity.

- Per interaction with the police, a black civilian is 1.5x more at risk to receive slight use of force by the police than a white civilian. This disparity gets smaller at higher levels of force. (See [7], Figure 1)

- Per interaction with the police, a perfectly compliant black civilian is 1.2x more at risk to receive slight use of force by the police than a perfectly compliant white civilian. This disparity gets smaller at higher levels of force. (See [7], Figure 5)

[1] https://www.pnas.org/doi/10.1073/pnas.1903856116

[2] https://www.pnas.org/doi/10.1073/pnas.1919418117

[3] https://www.pnas.org/doi/10.1073/pnas.1920184117

[4] https://twitter.com/gradempunion/status/1270829004157849600

[5] https://lansingstatejournal.com/story/news/2020/06/19/msu-vp...

[6] https://www.pnas.org/doi/10.1073/pnas.2004734117

[7] https://scholar.harvard.edu/sites/scholar.harvard.edu/files/...

Wow, this is really interesting. Per interaction is doing a lot of work here. Perfectly compliant seems pretty flexible as well.

Bayes’ rule and conditional probabilities have definitions.

Reply on news.ycombinator.com