How could one design a test for the last statement above (which resembles a hypothesis)? In practice it seems the statement "varies greatly depending on how the

How could one design a test for the last statement above (which resembles a hypothesis)? In practice it seems the statement "varies greatly depending on how the individual uses it" equates too easily with "varies greatly depending on the individual" (fundamental attribution error). Also, it may be that we believe we are very good at using something to our advantage when in fact we should not be using it at all (casinos exploit this problem). The reality is that it is impossible to split test oneself to discover whether in fact there was a benefit, although I wish this were possible. Maybe try something very difficult in two scenarios: play a game of skill which may benefit from looking something up (like chess openers) several times without a phone and several times with access to the phone to see whether there is a difference in performance against a continuously adaptive algorithmic opponent.