I didn't claim the participants wouldn't know whether they own a phone - obviously they would. I said they wouldn’t know they’re in a study whose purpose is to correlate smartphone use with academic performance.
That's perfectly compatible with a double-blind setup:
* the *students* just think they’re taking standardized tests, not that the effects of their smartphone habits are being monitored;
* the *graders* don’t know whose tests belong to whom.
That’s about as "double blind" as social-science research gets. The commenter I replied to latched onto the literal impossibility of hiding the phone itself, not the intentional design of the experiment.
