We Can't Improve Behavior by Fudging Discipline Stats
In this video, Dr. Justin Baeder argues that manipulating discipline data to look better on paper doesn't actually make schools safer.
Key Takeaways
- Data manipulation is widespread - Schools under pressure to reduce suspensions often stop reporting incidents rather than actually reducing them
- Fake data produces fake progress - Declining suspension numbers that don't reflect real improvement create a dangerous illusion of safety
- Honest data drives real improvement - Only when schools accurately track behavior can they identify and address actual problems
Transcript
You can't make a school safer by cheating on your discipline statistics.
And that should be obvious, but I see a lot of schools and districts and even whole states trying to influence statistics in thinking that that'll have a magical effect on reality.
And of course we should measure what matters, but we have to understand Campbell's Law that when you attach accountability pressure to a measurement, you inevitably end up distorting the process that you're intending to measure.
For example, a lot of states have banned out-of-school suspension for almost all offenses for students in grades K-3.
And we might want to improve outcomes in those areas, but there's no magic happening.
And hopefully people are starting to realize that you can't take one tool away without having a good replacement in place.
And a lot of the promises that have been made about PBIS and restorative practices and other alternatives to suspension have turned out to not really work.
And that's a problem by itself.
But I also don't want to downplay the outright cheating on discipline statistics that takes place in the form of downgrading behaviors or simply failing to report behaviors because they would make us look bad.
If we want to improve in a particular area, yes, we need to measure what matters, but we have to measure accurately.
We have to measure honestly.
And if we're constantly fudging the numbers and reclassifying things to make ourselves look better, then that is absolutely not going to lead to improvement.
It's going to lead to improvement on paper, but not in the reality for teachers and students.
The NYPD famously got caught at this.
They had a system called CompStat where captains were held accountable for crime in their district.
And of course what happened is that crimes would simply get downgraded or not reported at all.
And when you do that, your statistics look better, you look better, but the reality does not improve for anyone.
So when an assault gets reported as a scuffle or a fight, or when bullying gets downgraded to just a conflict, we're doing the same kind of thing in education.
If we want to get better in all of these areas we have to have honest statistics.