Originally Posted by svenax
Surely in one you are talking about the presence or absence of a phone (presumably switched on) while the subject is sleeping without making any particular claims about whether it's heat, signal, whatever... causing the problem.
In the other experiment, they aparantly want to test whether wifi stops seeds germinating. To do that the way they've done it, they need to control for any other effect the computers and routers may be having on the seeds other than wifi. Maybe the best way to do that would be to run the experiment again having disabled the wifi.
Wouldn't it have been easier to put a phone in a bag and sleep next to it without knowing whether it was on or off? You could even do the traditional verification of making sure the effect you are looking for is there with the experiment unblinded, and then blinding the experiment.