In one famous study (Rosenhan, 1973), eight healthy volunteers ("pseudopatients") presented themselves to psychiatric hospitals. They told the admissions officers that they had been hearing voices. Except for this one lie, they answered all other questions truthfully. All pseudopatients were admitted to the hospitals, where they ceased to pretend they were hearing voices (in other words, they behaved ‘normally'). None of the pseudopatients were detected as sane. Rosenhan argued that the staff at the hospitals seemed to be operating under the null hypothesis that patients in a psychiatric hospital are insane.
In this study, did the hospital staff commit a Type I or Type II error in calling sane patients insane? Which would be the more costly error? Why?