Orne and Holland (1968) research
criticised Milgram's low of internal validity in his baseline study
they point out that his variations had extra manipulation of variables
an example is the variation where the experimenter is replaced by a ‘member of the public’- Milgram recognised that this situation was quite obviously fake