Mass Shootings

Federal Survey Wildly Overestimates Number of School Shootings

The Department of Education's data on school shootings is riddled with inaccuracies.

|

Soleg1974/Dreamstime.com

The federal government has no idea how many school shootings occur each year.

The Department of Education (DOE) says that there were 235 school shootings in the 2015–16 school year. But when NPR double-checked the figure, it found that only 11 could be independently confirmed. A full 161 of the reported shootings—some two-thirds of the alleged total—did not happen at all. Another four incidents were miscategorized. (In one case, for example, a cap gun was fired on a school bus. In another, a student posted a picture online of himself holding a gun, at home, on a Saturday.) And 59 could neither be confirmed or disconfirmed, since the districts never responded to NPR's inquiries.

The DOE's figures appeared in 2018's Civil Rights Data Collection (CRDC) survey, a biennial report that asks districts about everything from school disciplinary rates to the availability of advanced placement courses. This was the first time the survey asked schools if they'd had any shootings.

At the very least, this shows that a lot of time and effort was spent collecting useless data. At worst, this raises the possibility that these bad stats will be used to direct millions in federal school safety grants toward unproductive or even counterproductive uses.

Ventura Unified School District in Southern California reported 27 school shootings to the DOE but told NPR that the correct number was zero. An assistant superintendent speculated that someone had hit the wrong button by mistake.

Cleveland's school district reported 37 shooting incidents to the DOE, then told NPR that this was actually the number of times people had been caught on school grounds with a firearm or a knife. Someone had accidentally entered the number on the wrong line.

The DOE tells NPR that all this data will have to stay in the system, as the department's deadline for corrections has closed. But the department will amend its latest survey with a note about the problems NPR uncovered.

I hope they notice that note at the National Institute for Justice's Comprehensive School Safety Initiative (CSSI), which is designing pilot programs to cut down on school violence. CSSI is relying on that CRDC data to evaluate the trends in school shootings. Obviously, any analysis based on these numbers is going to be flawed. (In March Congress appropriated $75 million for the CSSI's efforts.)

Past programs crafted off CRDC data have proven disappointing, even when the data may have been more sound. An Obama-era effort to combat racial bias by intervening in districts that were disciplining black students at higher rates, for example, prompted some schools to game the numbers. School safety is an innately local problem, and the best ways to address it will be crafted by people with local knowledge, not by federal bureaucrats tasked with interpreting numbers from faraway districts. Especially when there's a good chance those numbers aren't remotely accurate in the first place.

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

87 responses to “Federal Survey Wildly Overestimates Number of School Shootings

  1. Well, it’s about guns, so who needs facts?

    1. The DOE tells NPR that all this data will have to stay in the system, as the department’s deadline for corrections has closed.

      “Look, we know the data is catastrophically wrong, but it’s just too late to do anything about it, we’ll just have to deliver these wildly inflated numbers to the people making the decisions. Oopsie!”

    2. Even if there weren’t that many shootings, that’s only in the past. The number of future shootings is bound to skyrocket.

      1. *tsk tsk tsk*

        Proper punning snark technique would have led you to the answer, “The number of future shootings is bound to shoot up” or “In the future, these numbers will be on target“. 😉

      2. Ipse dixit.

    3. No, it is about grant money and how to increase the pot of money.

  2. I was amazed to see this on NPR Morning Edition today. It’s almost like reading that the ACLU is supporting the NRA in NY. Poe Horizon; we can’t tell “real” news from The Onion.

    1. NPR has gone off the reservation.

      1. I’m sure the that reporter will be disciplined.

      2. The intern responsible has been reassigned to a nice new Gulag in Siberia. All is well comrades.

    2. Was it the bad numbers or the correction that was on the news.
      I can’t blame them too much. I would have taken the DOE’s word on it as well if it was an official publication.

      For something so important as a school shooting, you would think that they would ask for a list of all incidents including dates, number injured, and number killed.

    3. Actually the ACLU had similar findings, they just opted to bury the information in their reporting, and not make a fuss over it.

  3. “…An assistant superintendent speculated that someone had hit the wrong button by mistake.”

    ?

    Is it me or this needs more ‘beef’? As in details.

    1. Ventura Unified School District in Southern California reported 27 school shootings to the DOE but told NPR that the correct number was zero. An assistant superintendent speculated that someone had hit the wrong button by mistake.

      “Possibly someone hit the ’27’ button instead of ‘0’.”

      1. Or maybe they hit the +1 button. You know, 27 times. By mistake.

        1. “The clerk was understandably quite nervous, having to report to Washington.”

          1. She’s (he’s?) no dummy. She’s read the Clinton Inc. body count.

          2. Betsy DeVos just makes everybody so jumpy they just start hitting wrong buttons left and right. Trump’s fault for appointing her.

      2. Are you saying that the top row of your keyboard doesn’t read ” ` 1 2 3 4 5 6 7 8 9 0 27 – = backspace”?

        What sort of weird ass keyboard do you have?

      3. I read that statement as a “I think it’s an entry on wrong line but I have no idea” statement.

      4. Its like a series of tubes.

        1. I laughed.

    2. Is it me or this needs more ‘beef’? As in details.

      His secretary was supposed to have incorrectly connected him to the call so that he could pretend he was unavailable to answer questions, she accidentally hit the wrong button and connected him by mistake.

    3. Most likely Ventura hired the clerk from Hawaii who “accidentally” triggered the missile alert in January by also hitting the wrong button. Apparently it’s the new “go to” explanation, having overtaken “my dog ate it”

  4. The DOE tells NPR that all this data will have to stay in the system, as the department’s deadline for corrections has closed.

    The DOE has strict time requirements on what it considers facts. Facts that are submitted past the deadline will not be considered facts.

    #FactBased

    1. If you’d listened to the NPR story, they publish these statistics dead tree-style. They accept corrections up to a reasonable time before it goes off to the printer. After which, it’s published with errata.

      Also, yeah, six months after a consensual sexual encounter it doesn’t become a rape any more than a rape becomes a consensual sexual encounter 6 mos. after the fact. They should check their facts but the DOE is not a criminal investigative organization.

      1. it is both understandable and frustrating.

      2. After which, it’s published with errata.

        Errata:
        Step 1: Starting on Page 1, Line 1, Strike out “235 school shootings” and substitute “11 school shootings”
        Step 2: Repeat correction 23,476 times.
        Step 3: Drop report in Recycle Bin.

        I wonder how many calculations in the report, on how many tables, depended on “TotalSchoolShootings = 235” as a variable.

  5. “…Past programs crafted off CRDC data have proven disappointing, even when the data may have been more sound. An Obama-era effort to combat racial bias by intervening in districts that were discipling black students at higher rates, for example, prompted some schools to game the numbers.”

    I would love to know more about how this turned out in practice and if it helped black students at all.

    I’m guessing lotsa unintended consequences.

    1. No unintended consequences. The intention was to help Emperor Hussein, not black students.

    2. Like the kid at M.S. Douglas H.S.?

  6. Credit where due. Media checking government.

    MOAR PLZ!

    1. NPR no less. Very impressed. They have their moments.

        1. a dead clock is right twice a day…

          1. Not if it’s digital

          2. IMO, the problem with NPR isn’t that they’re never right as much as they’ll (metaphorically speaking) tell the approximate right/relevant time maybe ~33% of the time, another ~33% of the time they’ll report the time as 78:34 a.m., and ~33% of the time they’ll tell you it’s morning because their correspondent in Shanghai says it’s the morning there.

            Now, sure, most of the time they tell you the correct time it’s at a time or on a topic where no one’s listening but they’re correct nonetheless.

        2. National Progressive Radio correcting wildly misleading firearms stats?

          It’s got to be a trap.

  7. The Department of Education doesn’t do anything useful and should be closed down.

    Cue all the assclowns who will scream “but then nobody will know how to read!” Haven’t they ever noticed that people learned how to read long before there was such a thing as a “Department of Education”?

    1. “But those people *owned slaves*!”

    2. It’s a fact, no one could read before 1979.

    3. The fact that now, with the DoE, high schools are graduating people who can’t read, is irrelevant.

      “But they have a Diploma!

  8. The DOE tells NPR that all this data will have to stay in the system, as the department’s deadline for corrections has closed.

    Over/under on how often this “data” will be cited by left-wing yakkers and bloggers?

    1. Only some crazy right wing nut job doubts official US government data, duh!

      1. “In other news today, noted alt-right broadcaster NPR was booted from Facebook after questioning official…”

        1. Nope, too accurate.

          “after engaging in hate speech…” natch.

  9. The DOE tells NPR that all this data will have to stay in the system, as the department’s deadline for corrections has closed.

    “Hi. We’re from ’60 Minutes’. We’d like to speak with your supervisor.”

  10. >>>Past programs crafted off CRDC data have proven disappointing,

    “civil rights data collection” wins stupidest thing I’ve read today.

  11. This is actually unavoidable when measuring rare events. Mistakes in data gathering invariably inflate the rare event than deflate it. They hinted at this in the NPR piece this morning.

    There were 96,000 schools reporting in this survey. If something happens 1 time in 10,000 — e.g., 11 in 96,000 — then a mere 2 in 1000 error rate will make it look like 200 times in 96,000. Using those numbers, 0.2% of the 95,089 will incorrectly report that the rare event occurred, while 0.2% of the 11 — i.e., none — will incorrectly report that the rare event did not occur.

    That the people compiling these surveys don’t recognize this should be shocking.

    Ha. Just kidding. It’s the government: it should be expected.

    1. Ha. Just kidding. It’s the government statistics: it should be expected.

      All but the most empirical use of statistics on non-human data should come with a little Intel-knockoff logo that says “State Inside”.

  12. Dark state indeed.

  13. Deep state

  14. Also, what does take to get a ‘hat tip’ around these parts? Have I set my expectations too high with Britschgi.

    1. Always remember: It was Britches that destroyed PM Linques.

      Bastard.

  15. an article without a ‘to be sure, even one shooting is too many’ is the best part of the day

  16. >>>The DOE tells NPR that all this data will have to stay in the system, as the department’s deadline for corrections has closed.

    There’s a statute of limitations on accuracy.

    1. It’s when the truth isn’t the truth.

    2. Missing a corrections deadline is how alternative facts are born, it seems.

  17. Don’t worry, I am sure that the progs are hard at work to create a narrative where:

    1. Someone in the Trump administration ordered the DOE to tamper with the study.
    2. Someone in the Trump administration ordered NPR to fabricate the story,
    3. Russian hackers have taken over NPR.

    Take your pick

    1. “”Take your pick””

      I’ll wait for the book.

    2. Why do you assume that the story will contain only one of those elements? 😉

  18. So how many listeners did NPR lose, who will now only get their news from HuffPo, Salon, and MSNBC, because of this?

    1. Publicly funded radio. They don’t care about listener numbers (and it shows) they only care about donor numbers.

      A couple more investigative stories like this and I might donate. Especially because on the radio they teased that it was largely the work of unpaid interns that contacted and clarified things with the reporting districts. I certainly feel like less of a libertarian when NPR, of all places, is using their orphans to for something constructive like sniffing out and exposing bogus government statistics rather than work the salt mines.

  19. The picture would be more appropriate if the first crossed out number was “5”.

    But I suppose they would’ve had to go to Soleg1984/Dreamstime for that.

  20. Well, yeah.
    The federal government is going to “overestimate” gun shootings.
    How else are they going to justify spending billions of dollars on the public school monopoly?

  21. Carry on, clingers! I’m going to make some popped corn to eat while I sit here and smirk at your feeble attempts to reply to this internet comment forum, wingnuts!

    *gets up carefully to avoid accidental bowel evacuation*

    1. *gets up carefully to avoid accidental bowel evacuation*

      That would be no accident. It’s just your creative process churning out another final draft.

    2. Rev Arthur L. Kirkland|8.28.18 @ 6:44PM|#
      “Carry on, clingers! I’m going to make some popped corn to eat while I sit here and smirk at your feeble attempts to reply to this internet comment forum, wingnuts!”

      A self-righteous lefty asshole smirking? Like Hihn needs company.
      Help yourself, loser. And fuck off.

    3. Dear Rev, a piece of advice: when you have nothing to say, the best course is usually not to say it

    4. Y’all so serious. Parody Reverend is much better than the real thing. The real Artie is so much more smug, and never attempts to be funny.

      1. I must admit, the prick (the real one) is such an insufferable ass that I just had to tee off on an obvious parody.

  22. “The Department of Education (DOE) says that there were 235 school shootings in the 2015?16 school year. But when NPR double-checked the figure, it found that only 11 could be independently confirmed.”

    Hey, 11 or 235? Good enough for government work!

    1. Hey, 11 or 235? Good enough for government work!

      Close only counts in horseshoes and hand gre… close enough for government work it is!

    2. 220, 221, whatever it takes.

  23. And no hihnfection under all those many fake accounts that piece of shit uses?
    Did we get lucky and have him/her carted off in an ambulance?

  24. Nice information. Thanks for sharing. Visit https://quickensupports.org/

  25. To control the people, control the statistics. People readily accept numbers.

    1. And “studies show….”

  26. Facebook took down a link to the NPR article because it looked like spam to them. Yup.

  27. When did a knife become a gun? All this survey proves is those opposed to guns have no problem altering data or flat out lying to make the problem seem much larger than it actually is to push their political agenda.

    1. And that’s not news. For decades, they’ve counted 24 year olds who lost a gang war as “children killed by guns”.

  28. Ask yourself this: Would I respond to inquiries from National Socialist Radio?

    1. Why not, if you work for the national socialist school system?

  29. Geek Squad Online Support provides online support for your damaged devices. The team of online support is experienced in dealing with tech issues online. In case you’re facing issue with your device contact Geek Squad Online Support

  30. I read your article in the head and stopped my work. All of them have got away with this information. And in the future, if any problem comes to me, then I will support your article and will advise others too. Once again I thank you for your day and your family will always be happy. https://bit.ly/2OCK214

Please to post comments

Comments are closed.