The school shootings that weren't.

[Carl N. Brown]

It was the first year that Department of Education asked that question of schools and school districts. The DoE should not have run with the first set of responses until there had been feedback on whether the question was understood and answered properly. Maybe the next year's survey with a properly vetted and widely comprehended clear question will get answered clearly.

NPR was able to confirm 11 actual schools reporting shooting incidents for the Fall 2015-Spring 2016 school year. Of the reported 240, 161 were mostly incorrect entries in the questionaire according to the schools. NPR did not get answers from 59 of the schools that allegedly reported shootings. (I suspect if there had been shootings the schools would have been quick to acknowledge that to a query by NPR. There would be more motive to not acknowledge an oopsie.)

For example, the DoE report stated 37 schools reported shootings in the Cleveland OH school district and 26 schools reported shootings in the Ventura County CA school district. Did you miss in the news media that there were 63 school shootings in just two school districts during the Fall 2015-Spring 2016 school year? According to NPR, Cleveland and Ventura County quickly claimed there were no shootings in their districts that year. Apparently the new question was inserted in a confusing position and the Cleveland and Ventura school districts owned up to the reporting error.

The sad thing is that it was NPR that did the follow up questions to the schools, when it should have been DoE verifying statistically impossible responses to a new question on their school survey.

As an aside: routinely in crime research like John Lott's last defensive gun survey, survey callers are instructed to call back later and see if the story about a positive response changes. In the NIJ NSPOF DGU survey, statistical anomalies (like a woman who claimed she used a gun defensively 52 times a year) were simply dropt without question. Does the DoE have the ability to do criminology research?

 

[CapnMac]

One of the things that is confusing (for wont of a better term) about NPR reporting is that they will report things as they are, rather than as they are expected to be.
Now, they do not change editorial direction because of any of these revelations, and this contributes to a perception that they are biased. Which they may well be; but, the evidence for or against that is largely subjective.

From my subjective point of view, I do not find their news product one which I desire. Are they as biased as Bill Mahr's VICE news? No (that's pure predetermined editorial content disguised as documentary). But, I just don't find NPR's reporting satisfying--which is likely from my subjective assessment of which observational biases they are unwilling to question.

My take-away on this is that this data is nice to have, particularly for the "cachet" than NPR puts upon it. As facts go, it's not the ace of spades; a jack of hearts that might take a trick, but nothing to bet a grad slam upon.