Deception (not fully informing a subject about the study as it might alter his/her behavior) must be justified by scientific necessity and benefit, and is assumed to be necessary only rarely.
Where deception is allowed, subjects must be thoroughly debriefed after intervention.
