That's only relevant if it affects one group more than the other, maybe you can control for this or take another approach, in any case I would find the results way more convincing than a survey.

Especially considering that going through a tough time will colour your memory and memory is extremely unreliable.