I read about at least one study (different description, illustrating the same principle) where they gave a set of teachers a list of students' names with numbers next to them. Apparently it was implied (or just not stated, which would be similar here) that the numbers were test scores or IQs. Sure enough, there was a strong correlation between the number by the student's name and how well s/he tested after a year.
The numbers were random, of course. Specifically, they were locker numbers.
Locker numbers aren't entirely randoom - while it seems unlikely there'd be a correlation, I'm surprised they didn't just use real random numbers to rule that out.
Maybe walking further between classes makes people more likely to be late and less prepared to pay attention?
The numbers were random, of course. Specifically, they were locker numbers.