How reliable do you think it is to use participants as observers of their own behavior in interventions? In what situations could this be helpful?

How reliable do you think it is to use participants as observers of their own behavior in interventions? In what situations could this be helpful?
Post Opinion