Comment by rahimnathwani

This criticism would be valid if the researchers had studied just one group of schools, and their methodology was just comparing before and after. But that's not all they did. They had two groups of schools, with low/high cell phone use before the ban. Their hypothesis was that the schools with high cell phone use would see a larger change in test scores (as they would have the largest drop in mobile phone usage).

  We then turn to our causal analysis comparing schools with different degrees of apparent pre-ban student cellphone use, after vs. before Florida’s cellphone ban. We show that the ban increased disciplinary incidents and suspensions significantly in the first year, immediately after the district started referring students for disciplinary action for cellphone use infractions. In particular, our difference-in-differences estimates suggest that the ban increased suspension rates by 12 percent (relative to the comparison group mean) and in-school suspension rates by roughly 20 percent in the first year.

There may be other reasons to criticize the paper, of course.