Confederates stood in the street and asked passers-by to perform tasks (picking up litter etc)
People were twice as likely to obey the assistant dressed in the guard uniform than the jacket and tie
Cross-cultural replications
Meeus and Raaijmakers - participants were ordered to say stressful things in an interview to someone (a confederate) desperate for a job
90% obeyed
They also replicated Milgram's proximity findings - obedience decreased dramatically when the person giving orders was not present
Counterpoint to cross-cultural replications (Meeus and Raaijmakers)
Smith and Bond identified just two replications that took place between 1968 and 1985 that took place in countries that were culturally very different from the USA (India and Jordan). Whereas the other countries involved were all very culturally similar to the USA (Spain, Scotland etc)
Therefore it may not be appropriate to conclude that Milgram's findings apply to people in all or most cultures
Low internal validity
Participants may have been aware that the procedure was faked
Orne and Holland - it is even more likely in his variation because of the extra manipulation of variables
Significant in the uniform variation - hard to believe