Jeff Asher explains a statistical problem!

WEDNESDAY, DECEMBER 18, 2019

Drum seems to mishandle Naep: We were stunned by something we saw in this morning's New York Times.

Online, the report appears beneath the headlines shown below. Incredibly, the writer turned out to be statistically competent. Could he possibly work for the Times?

South Bend and St. Louis, Where Crime Statistics Can Mislead
City limits can affect perceptions, and a criticism of Mayor Pete Buttigieg doesn’t hold up.

Say what? The New York Times was presenting an analysis of a potentially misleading type of crime statistic? It was even reporting that a certain criticism of Buttigieg's performance as mayor actually doesn't hold up?

Inevitably, it turns out that this report's statistically competent author doesn't seem to work for the Times. Online, the Times tells us this about that:

Jeff Asher is a crime analyst based in New Orleans and co-founder of AH Datalytics. You can follow him on Twitter at @Crimealytics.

As it seems to have turned out, Asher, being statistically savvy, doesn't work for the Times. As suggested above, we'd received that impression based on the apparent competence of his statistical work.

Asher's report concerns a potentially misleading aspect of the way big-city crime statistics can be compiled and reported. After Asher explains the statistical problem, a certain complaint about South Bend's murder rate under Mayor Pete pretty much disappears.

(Along the way, in the hard-copy Times, we were told to consult "the accompanying table"—an accompanying table which only appears online. Most likely, that error tracks to an editor who does work for the Times.)

As it turns out, the murder rate under Mayor Pete likely hasn't been as horrible as has been reported. We were happy to see this type of savvy statistical work because of a recent disappointment along statistical lines.

We refer to Kevin Drum's recent handling, or more likely his recent mishandling, of Grade 12 test score data from the Naep. Our favorite blogger put the data to traditional gloomy use in this recent post.

Absolutely nothing has worked in the public schools! In this case, Drum offered a variation on this familiar theme. He said that things have been working through Grade 8, but the gains are all gone by Grade 12.

It seems to us that we went over this statistical matter, in major detail, several years ago. We'll return to the topic tomorrow—but truly, when it comes to the public schools, the insistence on gloom, like Neil Young's rust, simply refuses to sleep.