3 ms·
School privatization has some wicked confounds for any claims of success because it's almost impossible to do a true randomized trial. In many cases, the report
by acdha 11y ago
School privatization has some wicked confounds for any claims of success because it's almost impossible to do a true randomized trial. In many cases, the reported success is due to cherry-picking the best students; in others, the lead disappears over time as regression to the mean takes effect as a school's population grows large enough to average out a few stellar performers. The really positive news coverage almost universally neglects to make any attempt to do a demographically-matched comparison and in most cases the effect disappears once you control for poverty, family stability, etc.
This isn't to say that charter schools are cheating, only that something this complicated is easy to misunderstand. It's also important to look at other trade offs: perhaps a school's performance is actually due entirely to attracting students who would be successful almost anywhere – a city is almost certainly better off if they stay there rather than moving to a nearby suburb, even if that doesn't directly help with the question of how to help disadvantaged kids.
- KMag 11y agoSet up a double-blind test, where there's a fake private school where the principal on down don't know that they're actually government-run. At the end of the school year, have Aston Kutcher show up to the school with a TV crew and tell everyone that they've been working at/attending a public school for the past year!