4 ms·
The Stanford data set suffers from the same problems as the one in discussion: It opens up for correlating 'race' with foul behavior. Of course there may be suc
by stareatgoats 8y ago
The Stanford data set suffers from the same problems as the one in discussion: It opens up for correlating 'race' with foul behavior. Of course there may be such a correlation per se, just as there may be a correlation between heavy drinkers/medication abusers and excessive use of force. The problem with isolating the 'race' factor is that it promotes race hate, which is just stupid, when there are factors that are amendable ('race ' is not amendable). There is obviously a correlation between racism and inappropriate police work, but these kinds of data sets don't show that.
> I wonder how you propose that a database "analyze racism"?
I note that we are in agreement: you can't surmise racism from this kind of data, still that's what the presentation implies. To be able to establish the extent of racism one would have to conduct attitude research (psychological profiling) on the officers.
- jakelazaroff 8y agoYou're not in agreement. GP is saying that "racism" isn't a quantifiable metric, so you have to deduce it from data that include race. You're saying that it's not "racism" without officer intent, which is entirely semantic; data can still show racial discrimination.
- Falling3 8y ago>The problem with isolating the 'race' factor is that it promotes race hate How is your stance any different from the common and absurd refrain that discussing racism is racist? We can't discuss racism without looking at race.
- stareatgoats 8y agoI'm not familiar with that refrain, and agree it sounds absurd. FWIW I really do believe we need to discuss racism, however not with racist arguments.
- Falling3 8y agoI don't think anyone here will disagree with you on that either. But I fail to see the relevance of that in the context of this article.
- danso 8y agoYes, data can be misinterpreted and abused. But you seem to throw out the possibility that data can be correctly interpreted to cast doubt on previously held notions. The prevailing belief is that white cops punish non-white suspects. Data frequently shows that this is either not the case, or is significantly more complicated (e.g. similar numbers of non-white cops being involved in incidents).