Informed consent can't be fully utilised in psychological experiments, because knowning the true meaning of the test would destroy the results. With a purely physical procedure it's easy to explain to the subject what is going to happen - 'we're going to stop you sleeping for 48 hours and monitor your reaction time.' Here the test subject knows what will be done, but he won't change his behavior.

However with a psychological experiment - 'we're going to place you into a stressful situation and see how long it takes you to display aggressive behavior' - having the subject know will affect the result. Here the subject will be consciously aware that his aggression is being monitored, and so may make additional effort to restrain himself.

The solution is to not explain to the subject what the true aim of the experiment is, revealing it only after the experiment is complete. However this soon results in a subject showing combative behavior - one person who took part in several experiments said 'After the first few, it became a case of me against them. I'd being trying to work out what they were really measuring, and this would make me behave differently.'