++Research support for NSI. Asch (1951) found that many participants conformed because they were afraid of disapproval. When participants wrote down answers (no normative pressure) conformity fell to 12.5%. Shows that at least some conformity is due to a desire not to be rejected by the group for disagreeing.
++Research support for ISI. Lucas et al. (2006) found that participants conformed more to incorrect answers when maths problems were difficult. For hard questions, situation was ambiguous, so relied on (incorrect) answers given. This supports ISI because the results are what ISI would predict.
--Individual differences in NSI. Some people concerned about being liked by others - nAffiliators who have a strong need for 'affiliation' (need to relate to others). McGhee and Teevan (1967) found students who were nAfilliators were more likely to conform. Shows NSI underlies conformity for some more than others - an individual difference not explained by a theory of situational pressures.
--The NSI/ISI distinction isn't useful. Lucas et al.'s study shows that the NSI/ISI distinction may not be useful because it is impossible to work out which is operating.