2. Classic study - Rosenhan

On being sane in insane places

What was the purpose of the study?

• Firstly, he wanted to challenge the diagnostic system for mental health. 

• He felt that the diagnostic system  in place in the 1970’s emphasises the individual as the source of symptoms and ignored the environment in which the symptoms arose. 

• He also strongly believed that diagnosis was invalid as it was affected by observer bias. 

• He believed that clinicians diagnosed disorders by focusing on the symptoms that arise as a result of an underlying disorder (and simply matching these to a diagnostic criteria) rather  than in a certain context and judging them in accordance with this. 

What were the aims?

• To investigate whether sane people could be admitted to a psychiatric institution and once there, would they be detected.

• (To investigate the reliability and validity of diagnosing) 

• He also wanted to investigate what life inside a psychiatric unit was like and from this, raise awareness of such conditions. 

What design

What was the IV

What was the DV

What was the design 

What was the method 

• Design - Field experiment 

• IV - Symptoms

• DV - staff responses

• Design - ethnographic

• Method - Participant observation

Who were the pseudo patients?

• 8 pseudo patients 

• 3 women 5 men 

• Rosenhan and 7 others

• Mixed gender, background and age 

• Psychology graduate, painter, housewife, paediatrician, 3 psychiatrists and Rosenhan

Who were the participants?

• Staff and real patients at the institutions accessed.

How many hospitals were included? 

• 12 - representing a variety of institutions across 5 states.

• Some old, some new

• Some fully staffed, some understaffed

• Large and small

• Private and public

• Rosenhan attempted to have a varying amount of hospitals, so the diagnosis patients received from those hospitals can be representative of the diagnosis they would have received in any hospital in America.

• However, when looking at the size of America and size of institutions, the number is not as representative as Rosenhan thought. 

How did the pseudo patients access the institutions?

• They called up and asked to speak to admissions because they were ‘hearing voices’ they then attended the institutions and were admitted.

• Auditory hallucination - hollow empty thud

What were they diagnosed with? 

• All except one were diagnosed with schizophrenia - one with manic depression with psychosis. 

What were patients required to do once admitted?

• Act ‘normal’, record information and take notes about what was going on in the mental institution. 

• Convince staff that they were mentally sane. 

How was ‘normal’ behaviour assured?

• Visitors attended daily who confirmed ‘normal’ behaviour. 

What wasn’t detected by staff and what was the duration on stays in the institution?

• Sanity was not detected

• Minimum stay was 7 days

• Longest was 52/ a month

• Average was 19 days.

How did the real patients react to the pseudo patients? 

• They did detect that they were sane and some even questioned the pseudo patients. 

• 35 of a total 118 patients voiced their suspicions. 

How did the staff respond to the pseudo patients?

• They treated them in accordance with their diagnosis and pathologised normal behaviour.

• Staff and patients kept separate.

• Staff did not engage with real patients or pseudo patients.

• Some staff saw mental illness as something they could catch so they stayed separate.

• Physicians were less available during the day - were rarely seen on the wards.

How were the patients depersonalised?

• They were ignored, poor eye contact and a lack of verbal response. 

• Even worse with senior staff - 185 questions asked - non answered. 

• ‘Note taking behaviour’ 

How was unreliability of diagnosis confirmed at follow up?

• Rosenhan worked with one hospital and agreed to send in fake patients for their staff to try and identify.

• Over 3 months, 193 patients were admitted - 41 were thought to be fake by 1 member of staff - 19 were thought to be fake by 2 members of staff.

• All patients were real

• Type 2 error.

Why was Rosenhan concerned about labelling people through diagnosis?

• People then treat people in accordance with their label - perception of that person was changed because of the context they were in. 

• Leads to stigmas, stereotypes, prejudice.

What is the danger of labelling here? 

• Can lead to self fulfilling prophecies, if all staff are treating you in a certain way, and you are being labelled as mentally ill you could start to believe this and debate that you may be  acting abnormally and potentially play up to the label. 

Generalisability - A03

A01:

• Rosenhan used a varied sample of hospitals throughout America - 12 hospitals across 5 states.

• Some were well staffed, some poorly, some new, some old, one private, a research orientated hospital, federal and university funded hospital.

A03:

• The sample can be considered a strength of the research as generalisation is possible across all hospitals in America.

• The sample of hospitals can be considered as representative as it could be considered that other hospitals would have come to the same diagnosis.

• However, due the fact that mental health has significantly progressed since the 1970’s, it lowers the ability to generalise the findings from Rosenhan’s study to mental health clinical assessments in the present day.

• At the time of Rosenhan, the DSM11 was being used, however clinicians now use the DSM5 which should reflect the increased understanding of mental health over time and should be more reliable as diagnosing mental illness.

• Furthermore, the research was conducted in America only, which further limits the generalisability to other parts of the world, therefore the sample and results are unrepresentative of other countries who may use the ICD diagnostic manual instead.

Reliability - A03

A01:

• Rosenhan trained his pseudo patients to use a standardised procedure.

• All pseudo patients were told to report 1 common symptom of hearing a loud empty thud.

• Apart from the common symptom, they should tell the truth about their behaviour and experiences.

• All were told to phone the hospitals for clinical assessment and once they were admitted they should act normally.

• They were told to record the behaviour of staff inside the hospital during their stay.

A03:

• Using a standardised procedure can be considered a strength as Rosenhan’s experiment can be replicated in order to see if the results are reliable.

• Rosenhan or other researchers are able use the same procedure again to test whether a similar diagnosis would be received and if patients would be treated similarly once in the hospital.

• For example, in Rosenhan’s follow up study, misdiagnosis was again common throughout the hospitals used.

• This adds to the reliability of his results and calls into question the reliability and validity of mental health diagnosis at the time.

• However a limitation of the procedure was that not all pseudo patients followed the procedure properly.

• For example, one pseudo patient initiated a romantic relationship with one of the nurses.

• This possibly limits the reliability of the observations that were recorded due to a lack of focus and possibly bias or subjective opinions of the staff’s behaviour.

Application - A03

A01:

• Rosenhan aimed to find out whether sane people could be admitted to a psychaitric institution, and once there would they be detected - reliability and validity of diagnosing.

• He also wanted to find out how patients were treated once admitted into a mental health institution.

• He found that all pseudo patients were unreliably and invalidly diagnosed with a mental illness, and once admitted were further patholigised by hospital staff, ignored and treated with stigma.

• For example staff describing taking notes as ‘note taking behaviour.’

A03:

• His research has high application value as it had significant repercussions for the mental health diagnosis process.

• Due to the unreliable and invalid diagnosis, the DSM2 was revised to the DSM3 to try to address the issues and increase the reliability and validity of the diagnostic process.

• For example, there was further guidance on the exact symptoms and duration in order to lead to diagnosis.

• Clinicians also needed to be trained to take these changes into account and treat patients more holistically.

• In addition, hospital staff were trained on how to treat patients fairly and remove discrimination within the institution.

• Moreover the research by Rosenhan became part of psychiatric training.

Validity - A03

A01:

• Rosenhan used a field experiment for his research.

• Pseudo patients were part of the real mental health system, institutions and staff.

• Therefore the setting for this research was natural.

• The experiment was controlled by the IV - the symptom that pseudo patients reported experiencing.

• Some extraneous variables were also controlled.

A03:

• The setting is a strength as it has high ecological validity.

• Pseudo patients were diagnosed in the way that all patients are, and the behaviour of the institution staff was the natural behaviour they would display toward patients.

• This allows us to generalise the diagnosis and staff behaviour to hospitals outside of the experiment.

• Although Rosenhan was investigating the reliability of mental diagnosis, 11 out of the 12 hospitals misdiagnosed schizophrenia.

• This highlights that the pseudo patients were reliably misdiagnosed with an invalid mental illness.

• Furthermore, the fact that 1 pseudo patient was diagnosed with manic depression, highlights the unreliable diagnoses given, as they were not all consistently diagnosed even though they described the same symptoms.

• However it could be argued that the internal validity was as high as possible due to the fact that Rosenhan could control some extraneous factors, whilst achieving natural behaviour from the staff.

• On the other hand, the behaviour of some of the pseudo patients once in the hospital affected the validity of the patients and staff behaviour.

Ethics - A03

A01:

• Natural, participant, covert observation technique

A03:

• There are ethical issues as the real patients and hospital staff did not have the right to withdrawal or informed consent.

• This means that they were unable to withdraw their data, decide to withdraw from the experiment or give their consent in the first place.

• Furthermore informed consent cannot be gained from mentally ill patients as their judgement and understanding may be impaired.

• However the hospital management gave consent for the study to be done on behalf of the patients and staff.

• Rosenhan also completed a debrief, however this was only done with the hospital management.

• In addition the protection of the Pps (pseudo patients) was an issue due to the fact they did not know when they were allowed to be discharged from the hospital, which might have caused anxiety and may have caused them to act irrationally.

• Moreover, they all received a diagnosis of mental illness, which is a label that comes with prejudice and stigma from the hospital staff.

• When discharged from the hospital hey had a record of schizophrenia in remission - which means they were still regarded as having schizophrenia without present symptoms, which would still carry a stigma.