The 8 pseudopatients visited the 12 hospitals (some went to more than 1) and there were 3 females and 5 males. These included a psychology student, paediatrician, psychiatrist, painter, housewife and three psychologists plus himself. They all reported hearing an unfamiliar same-sex voice saying words of 'empty', 'hollow' and 'thud; which is not typically associated with SZ in their clinical interview and when they all got admitted, they behaved normally and observed.
RELIABILITY/VALIDITY
It was a covert participant observation as the staff wasn’t aware the pseudopatients were researchers. So, there is high ecological validity therefore the observed behaviour of staff would reflect how they treat their patients in everyday-life. An example is incidents of physical abuse occurred, staff member to patient, but when another staff member arrived, the abuse ceased. This quantitative and qualitative data collected by the researchers is valid as it was a naturalistic experiment.
However, Rosenhan selected people who supported him to be the pseudopatients with him so maybe they were biased in finding negative behaviours to back up Rosenhan’s predictions for treatment of patients.
Difficult to establish reliability as within each hospital, little data was collected, however comparison can allow for reliability if treatment of patients is similar throughout. Low inter-rater reliability within each hospital as many only had 1 researcher in them which means they could’ve easily missed behaviours.
The willingness to admit a patient on flimsy evidence might be down to them thinking that no one would self-admit who wasn’t mentally ill so they would have no reason to suspect someone pretending.