Sane in Insane Places


How can we define abnormality? What is meant by sanity? Who has the right to define what is normal? Is snacking on raw potatoes a cute oddity or simply abnormal? (One of the deep questions that plague my mind).

cyanide and happiness - crazy

The classification of abnormality is an area of psychology that constantly sparks debate. One way of looking at abnormality is through the “medical model”. The medical model has some basic beliefs, one of which is that all illnesses can be treated with medicine. Psychiatrists (cough-cough superiority complex alert cough-cough) are medical doctors who treat psychological illnesses similarly to how most physical ailments are treated.

In the 1950s, the medical approach – which we can also call psychiatry – used the DSM (Diagnostic and Statistical Manual of Mental Disorders) to classify abnormalities. A few years later in the 1960s, an anti-psychiatry movement grew. This movement criticised the medical approach to abnormalities as narrow and unreliable.

Rosenhan was one of these critics and decided to conduct an experiment that would lower the credibility of psychiatric classifications and prove their unreliability.

David Rosenhan, staring into your soul

David Rosenhan, staring into your soul

The sample in this experiment was made up of eight people: a psychology graduate student, three psychologists, a psychiatrist, a paediatrician, a painter and a housewife – three were females and five were men.

The participants were going to be “pseudo-patients”. They had to gain admission into 12 different psychiatric hospitals in five different American states. They telephoned the hospital, arranged an appointment and went to the admissions office. There, they complained about hearing voices. They said the voice was unfamiliar and unclear, and the same sex as themselves. They claimed they heard the voice saying things like “empty”, “hollow” and “thud”. These symptoms were partly chosen because of their similarity to existential symptoms that arise from concerns about life being meaningless (thinking things like Who am I and what is this all for?). Also, existential psychosis has not been previously mentioned in psychological literature.

The pseudo-patients gave the hospital a false name and job but all their other details were completely true, including their general life experiences, history and relationships. After they gained admission, they were told to stop showing any symptoms of an abnormality and convince the staff that they were sane. From here onwards, the pseudo-patients were responsible for getting themselves out of the psychiatric hospital.

The pseudo-patients then took part in activities and spoke to the other patients and staff members in an ordinary manner. When the staff asked them how they were feeling, they would respond by saying they felt fine and were no longer experiencing any abnormal symptoms. They also wrote detailed notes about the observations they made in the hospital. This was initially done very privately to avoid suspicion, but it eventually became more open because the staff was not bothered about any note-taking and interpreted it as the behaviour of an abnormal person.

In four hospitals, the pseudo-patients observed staff behaviour towards patients. Their observations were to show how a person would be treated when hospitalised in a psychiatric institution. The pseudo-patients would approach a staff member with a request and observe their reactions.

After this experiment:
Rosenhan’s experiment didn’t just stop after the pseudo-patients were released from their psychiatric hospitals. In a second part to his research, he carried out a second experiment. At a teaching and research hospital, the staff who were aware of the first experiment taking place where falsely told that new pseudo-patients would attempt to gain admission into their hospital next. Staff members were also asked to rate each new potential patient’s likelihood of being a pseudo-patient on a 10-point scale.

All of the pseudo-patients disliked this experience. They all wanted to leave the hospitals immediately but had to make an effort to be released (one participant stayed in the hospital for 52 days!). All of them were admitted into the hospitals with a diagnosis of schizophrenia and released with a diagnosis of “schizophrenia in remission” (basically non-severe schizophrenia). The hospital staff made this judgement without observing even one actual symptom of schizophrenia.

Although the pseudo-patients were not detected as imposters by the hospital staff, surprisingly, their fellow patients were more (correctly) suspicious. Out of 118 patients, 35 spoke out about this. For example: “You’re not crazy! You’re a journalist, or a professor. You’re checking up on the hospital.” This should probably make us question who is more qualified to diagnose sanity and insanity – the doctors or the “crazy” people.

One observation was that normal behaviour shown in the pseudo-patients was interpreted as a part of their perceived illness. The nursing records for three pseudo-patients labelled their writing as a factor of their pathological behaviour. Rosenhan noted here that there is a big overlap between sane and insane behaviour. Every person on earth must feel depressed at times, or maybe moody or angry or tearful on other days. However, in the setting of a psychiatric hospital, these everyday human emotions and behaviours were interpreted as a result of having a mental illness. Another example of this biased interpretation can be seen in an incident at a psychiatric hospital where a psychiatrist pointed to a group of patients waiting outside the cafeteria half an hour before the official lunchtime. This psychiatrist suggested that this behaviour was characteristic of an oral-acquisitive syndrome. A more likely explanation, however, is that the patients were probably just bored and had nothing else to look forward to.

The observations noted on staff behaviour towards patients were also compared with a similar study conducted in a university. The results were significantly different: the university professors acknowledged students and responded in a friendly manner whereas psychiatric patients were often ignored or dismissed.

In the psychiatric hospital, most patients were ignored or brushed off with a quick reply. Hardly any psychiatrists and pretty much none of the nurses stopped to talk to the patient properly.


Rosenhan described this experience as depersonalising for the pseudo-patients, making them feel very powerless. This was probably a result of the complete lack of consideration for the patients’ human rights, such as freedom of movement and privacy. For example: medical records were available for any staff member to open, most toilets did not have doors and personal hygiene was strictly monitored.  Some particularly idiotic staff members would be brutally violent or verbally abusive towards patients in front of other patients but would stop if other staff members approached, implying that the patients were not given any credibility.

The pseudo-patients were also given around 2,100 medical tablets. They only swallowed two; the rest were pocketed or flushed down a toilet. When the pseudo-patients went to the bathroom to get rid of their tablets, they usually found tablets of other patients already there. As long as patients were cooperative and avoided disturbances, this kind of behaviour went unnoticed.

Records were also made about the amount of time nurses spent with their patients. They noted that the nurses stayed in their offices for about 90% of their time. Also, the number of times staff came into the ward and the amount of time they spent with psychiatrists, psychologists and registrars was around an average of under seven minutes per day.

Results of the secondary study
The table below summarises the results of the secondary study. It illustrates how flimsy and subjective the labels of ‘insane’ and ‘sane’ can really be. After the hospital was told about the possibility of more fake patients, they incorrectly diagnosed mentally ill patients as clinically sane people.

The results here gave us two errors to consider: the Type 1 error and the Type 2 error. Type 1 error means to diagnose a mentally ill person as ‘healthy’, and a Type 2 error is to diagnose a ‘healthy’ person as mentally ill. The picture below should help you remember the difference.

type 1 and type 2 errors

Rosenhan used the results of both studies to support his hypotheses about mental illnesses and psychiatric labelling. He said that the first study showed psychiatrists’ failure to detect sanity while the second study showed the psychiatrists’ failure to detect insanity. However, Rosenhan wasn’t a meanie. He said that the failures of hospital staff should not be narrowed down to them having a bad or uncaring personality, but that the results were due to the environment and social stigmas. Rosenhan discussed how a more compassionate and personal hospital environment would benefit patients as well as staff.

Type of research method
This was a field experiment because it took place in a natural setting (the real psychiatric hospitals) but the researchers manipulated certain factors. Participant observation was also used by pseudo-patients in this experiment.

Independent variable
The fake symptoms of the pseudo-patients are the independent variable.

Dependent variable
The reaction of the staff to the fake symptoms (including their final diagnosis) and the treatment of the pseudo-patients can be seen as the dependent variable.


  1. Use of participant observation: By being participative observers, the pseudo-patients were able to remain as objective outsiders at the same time as experiencing incidents from a patient’s perspective. This makes their feelings, behaviours and reactions a little more realistic as they are looking at things from the patient’s role and are able to sympathise without losing all objectivity.
  2. Ecological validity: The study was a field experiment, meaning that it took place in a natural setting. This makes the results a lot more true to real life, rather than them being a result of an artificial situation.
  3. Quite generalisable: Rosenhan did not restrict this study to one hospital or even one state. He conducted the study in 12 different (some were well-funded, others were not, some were research-orientated whereas others were just shabby) hospitals in five different American states, making the final results more generalisable to American hospitals. If he had conducted the study in just one hospital, we would have been unable to apply the results anywhere other than that one hospital.


  1. Good and bad errors in diagnosis: Some researchers argued that Rosenhan was a bit too hard on the psychiatrists. Wouldn’t it be better to stay on the safe side and label a sane person as insane and treating them, instead of labelling an insane person as sane as letting them continue their life without treatment? Personally, this ‘weakness’ makes little sense to me. The bottom line is that wrong diagnoses were being given out, whether they were on the ‘safe side’ or not.
  2. New diagnostic criteria: When this study was conducted, the DSM-II was being used in hospitals. The DSM is now in its fifth edition and is said to have addressed many problems in diagnosis, thus making Rosenhan’s study slightly outdated.
  3. Some lack of ecological validity in participants: The pseudo-patients went through an unpleasant experience, but their emotions cannot be seen as exactly true to the real patients who did not have the comfort of knowing that their diagnosis was fake and that their stay was definitely temporary.
  4. Validity: The validity of the study is kind of questionable because the pseudo-patients reported hallucinations which are actually a very common symptom in schizophrenics. So, in a way, weren’t the psychiatrists diagnosing correctly at first based on the information they were given?

Ethical issues

  1. Informed consent: Rosenhan did not get consent from the hospital staff. They did not give permission to be used, or for their hospitals to be used, in any experiment.
  2. Deception: I think it’s clear to see that a whole lot of deception took place. The pseudo-patients deceived pretty much everyone in the hospitals about their true identity and the purpose of their admission.
  3. Confidentiality: Rosenhan kept the identities of the staff and hospitals private.
  4. Emotional or physical harm: There was no sign of any physical or emotional harm, although some of the pseudo-patients might’ve been a bit troubled by their experiences in the hospital. Also, it is possible that Rosenhan’s experiment may have prevented people with genuine mental illnesses from being treated because they may have been looked upon as imposters.
  5. The right to withdraw: Since the hospital staff and patients had no clue about the existence of an experiment, they weren’t really given the right to withdraw. They were not told about the experiment so they could not exercise their right to withdraw.
  6. Debriefing: There was no personal debriefing sessions.

Reference: Rosenhan, D. L. (1973). On Being Sane in Insane Places. Science. 179: 250-258.

3 thoughts on “Sane in Insane Places

  1. Hi!
    I just wanted to clarify one thing. The sample in the study by Rosenhan includes the staff members and doctors at the psychiatric ward. The pseudopatients are not the participants/sample because we conducted the study on the behaviours of the doctors, not the patients.


  2. Hi I just want to tell you about some few corrections in this
    1- all but one were diagnosed with schizophrenia,
    2- type 1 error is when you diagnose sane person as insane and type 2 is when you diagnose insane person as sane, (it’s written wrong in the books also written wrong in rosenhan study
    3- About the informed consent Rosenhan was the first pseudopatient and his
    involvement was known only to the hospital administrator and chief psychologist. So at least to the hospital he went to they knew about him

    • Hi Mashal,

      Thank you! This is really helpful and you’ve obviously got a good understanding of the study!

      Just wanted to note the following points though:
      1. The last participant was also diagnosed with schizophrenia BUT it was “in remission”.
      2. Yep, Rosenhan actually mixed up the definitions in his original research paper. However, candidates are never penalised for using his definitions as, of course, the syllabus references the original study. Luckily, both definitions are acceptable in the exams.
      3. Yes, Rosenhan told the “higher-ups” about the study but the actual test subjects (staff members and nurses, who were the targets for observation) were not asked for consent and this is relevant because it was these people who were actually part of the study itself.

      Again, thanks for going into detail about the points – it always helps to know the details and it’s great to see a student so aware! 🙂

      Best wishes,

Leave your comments below!

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s