About . . . . . . Classes . . . . . . Books . . . . . . Vita . . . . . . . Links. . . . . . Blog

by Peter Moskos

May 17, 2016

Let's all put our thinking caps, everybody

Why in the world are people reporting on last year's data when we have current data at our fingertips? Does reality only happen after numbers are published on DOJ letterhead?

NBC News reports:

Number of Police Officer Killings Drops, Reversing 2014 Spike.

Do reporters not have the wherewithal to see what the current situation is? It is but a mouse-click away. Officer Down Memorial Page keeps an excellent running and historical tab.

This year's data show killings are up (albeit with a small n of 17). We're on pace for 45 cops being shot and killed this year, and that would be much more than last year, and similar to the 47 shot and killed in 2014.

Now I wouldn't want the headline to say "Police Killings Spike 55 Percent in 2016," but at least that would be technically true. But it would be statistically irresponsible, with such a low n.

Oh, look what I just found. This is too good to be true:

2016 Sees Year-Over-Year Spike In Cops Killed By Gunfire.

And it's also from NBC News! (File under "You can't make this shit up!") Well, at least NBC has all their bases covered.

Anyway, it looks like 2015 is going to be the blip, not 2014.


[I corrected a wrong number I had regarding 2014.]

4 comments:

Andy D said...

Or maybe it won't be the blip, depending on how the year pans out. That small 'n' cuts both ways until it's over. Of course given that the killings are UP in 2016 why bother to write the first article?

David said...

Weird I don't see where they get +55%. If you look at the roster of 2015 non-accidental gunfire deaths, I count 13 by mid-May 2015. So far it's 17 this year, 31% higher than this time 2015.

Peter Moskos said...

I do not double check their work (I trust them enough, based on past experience.) But I've noticed they don't update the percentages continuously. Seems like they change one a week or so? So it might just be a week behind. And one or two deaths at this time last year And if It might just be a week behind. 6/11 gets you 55%. So it's probably using 11 for 2015 and 17 for this year. I bet it will change soon.

David said...

Yeah no I definitely trust them. It's a great (unavoidably morbid) resource for understanding what's going on in the world. Just trying to get to the bottom of the data...