You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
In 1973 the journal Science published a study that shook psychiatry. David Rosenhan's "On being sane in insane places" asked a question of startling simplicity — can the sane be reliably told apart from the insane? — and answered it by sending healthy volunteers, himself among them, into psychiatric hospitals pretending to hear a single word. Every one was admitted, most with a diagnosis of schizophrenia, and once inside not a single member of staff detected that they were, in fact, perfectly well. The study became one of the most cited and most controversial pieces of research in the history of clinical psychology, and it is the prescribed key research for the historical-context topic of the OCR Issues in Mental Health section. This lesson tells the story in the depth OCR's core-study format demands: the background and aim, the method (design, sample, procedure), the results, Rosenhan's conclusions, and a full, balanced evaluation. It is the empirical hinge of the whole section — everything you have learned about definitions, reliability and validity is put to the test here.
| This lesson covers | OCR H567 Component 03, Section A topic | AO focus |
|---|---|---|
| Aim, background and context of Rosenhan (1973) | Issues in mental health — historical context, KEY RESEARCH | AO1 knowledge of the study |
| Method: design, sample, procedure (Study 1 and Study 2) | Key research — On being sane in insane places | AO1; AO2 methodological understanding |
| Results: admission, labelling, "stickiness", depersonalisation | Key research — findings | AO1 |
| Conclusions and Rosenhan's argument | Key research — conclusions | AO1; AO2 |
| Evaluation: methods, ethics, validity, reliability, sampling, impact | Key research — evaluation; issues and debates | AO3 |
The specification is referenced descriptively throughout; consult the official OCR H567 specification document for the exact published wording. This lesson develops AO1 (detailed knowledge of the study), AO2 (understanding its methodology and applying its findings) and AO3 (evaluating its strengths, weaknesses, ethics and lasting significance). The full citation is Rosenhan, D. L. (1973) On being sane in insane places, Science, 179(4070), 250–258.
By the early 1970s, psychiatry was under intellectual siege. Critics in the "anti-psychiatry" movement — among them Thomas Szasz, whom you will study later — argued that psychiatric diagnoses were not the objective medical facts they claimed to be, but labels shaped by context, culture and the observer's expectations. The reliability and validity of diagnosis (the very issues covered in the previous lesson) were fiercely contested. Rosenhan, a psychologist and lawyer at Stanford, set out to test the claim empirically. If diagnosis were valid, then the "sane" and the "insane" should be distinguishable; the characteristics of a person should reside in the person, not in the situation they are in.
The intellectual context deserves a little more colour, because it explains why the study landed with such force. The 1960s and early 1970s were the high tide of the "anti-psychiatry" movement, a loose current of thinkers — Szasz in the United States, R. D. Laing in Britain, and the sociologist Erving Goffman among them — who in different ways questioned whether psychiatry was the neutral medical science it claimed to be. Goffman's influential work on "total institutions" had already described how asylums, prisons and similar closed settings systematically strip inmates of their ordinary identity and autonomy, remaking them into the passive role the institution expects. The labelling theorists in sociology argued that being labelled deviant could itself set off a process that produced further deviance, as the person and those around them lived up to the label. Rosenhan's study can be read as bringing these sociological ideas into psychology and putting them to an empirical test: if diagnosis were really the objective identification of an internal condition, the labels should stick to the genuinely ill and slide off the well; if instead the critics were right that labels and institutional contexts shape perception, then a sane person, once labelled and admitted, might find the label impossible to shed. That is precisely the hypothesis the pseudopatients were sent to test.
Rosenhan framed the central question as whether the salient characteristics that lead to psychiatric diagnoses reside in the patients themselves or in the environments and contexts in which observers find them. Put more sharply: can psychiatry reliably distinguish the sane from the insane? His aim was to find out whether normal, healthy individuals presenting with a minimal, ambiguous symptom would be recognised as sane — and, if admitted, how they would be perceived and treated inside the institution. The study therefore had two connected purposes: to test the validity of psychiatric diagnosis, and to describe the experience of hospitalisation for those labelled mentally ill.
The study is best understood as a field experiment combined with participant observation, using covert methods. It actually comprises several parts; OCR expects you to know the two central strands.
Sample. Rosenhan recruited eight sane pseudopatients (including himself). They were a varied group: three women and five men, comprising a psychology graduate student, three psychologists, a paediatrician, a psychiatrist, a painter and a homemaker. To avoid embarrassment to the institutions and to protect the pseudopatients' own careers, they used pseudonyms, and those in mental-health professions gave a false occupation (so staff could not treat them differently). Between them they gained admission to twelve different hospitals across five US states — a spread including old and new, well-funded and under-resourced, and one private hospital — so the findings could not be dismissed as peculiar to one poor institution.
Procedure. Each pseudopatient telephoned a hospital for an appointment and, on arrival, reported a single symptom: they said they had been hearing voices that were unfamiliar, of the same sex as themselves, and that seemed to say the words "empty", "hollow" and "thud". These particular words were chosen deliberately because they suggested an existential crisis — a concern that one's life is empty and hollow — and because there is no published literature reporting such an "existential psychosis", so the complaint could not be matched to any real condition. Apart from this single fabricated symptom and the false name and occupation, the pseudopatients gave a completely truthful account of their life history, relationships and circumstances — none of which contained anything pathological.
Crucially, the moment they were admitted, each pseudopatient stopped simulating any symptom whatsoever and behaved entirely normally. They said the voices had gone, cooperated with staff, and reported feeling fine. Their task from that point was simply to behave sanely and to observe — recording (at first covertly, then, as they realised no one cared, openly) life on the ward and how they were treated. They were told, as a genuine incentive to act convincingly sane, that they would have to secure their own discharge by convincing staff of their sanity.
Rosenhan then ran an ingenious follow-up at a teaching and research hospital whose staff had heard about the first study and insisted such errors could not happen with them. Rosenhan informed this hospital that, at some point over the following three months, one or more pseudopatients would attempt to gain admission, and asked staff to rate, on a scale, how confident they were that each new patient was a fake. In fact — and this is the sting — Rosenhan sent no pseudopatients at all.
Every one of the pseudopatients in Study 1 was admitted. In eleven of the twelve cases they were diagnosed with schizophrenia; in the one private hospital, with manic-depressive psychosis (a diagnosis carrying a more optimistic prognosis, which may reflect the different clientele). Not one was identified by staff as sane. Their length of stay ranged widely — from around 7 days to as long as 52 days — with an average of roughly 19 days before they secured release.
Perhaps the study's most damning finding is what happened to the diagnostic label. When the pseudopatients were finally discharged, they were not pronounced sane or misdiagnosed; they were recorded as having schizophrenia "in remission". In other words, the institution never concluded it had made a mistake — it concluded the illness was merely dormant. The label, once applied, was "sticky": it coloured everything and could not be shed. Rosenhan argued that a psychiatric label carries such power that it becomes a self-fulfilling frame — once someone is "a schizophrenic", their entirely normal behaviour is reinterpreted through the diagnosis.
He gives vivid examples of this reinterpretation. Pseudopatients openly wrote notes about ward life; staff recorded this note-taking itself as a symptom, one nurse writing that a patient "engages in writing behaviour". One pseudopatient's truthful, normal account of a close relationship with his mother that grew more distant in adolescence, and a warm relationship with his wife with occasional ordinary arguments, was rewritten in his case notes in pathological terms — as evidence of unstable, ambivalent relationships — because it was now being read through the lens of the diagnosis. A pseudopatient pacing a corridor out of boredom was asked by a nurse whether he was "nervous". The behaviour was normal; the label made it a symptom. Rosenhan called this the tendency to fit all evidence to the diagnosis: once the label is fixed, its "stickiness" distorts perception of everything the person does.
Although no member of staff detected the pseudopatients, a striking number of genuine patients did. In three admissions for which records were kept, many real patients voiced suspicions — "You're not crazy. You're a journalist, or a professor. You're checking up on the hospital." The people best placed to spot sanity were not the trained professionals but the fellow inmates.
The pseudopatients' observations of ward life exposed profound depersonalisation — patients treated as though they were not fully present or fully human. Staff, especially higher-status ones (psychiatrists more than nurses, nurses more than attendants), spent very little time on the wards and avoided contact. In one measure, pseudopatients approached staff with a polite, normal question (such as "When am I likely to be eligible for grounds privileges?"). Psychiatrists and nurses frequently failed to respond at all, or moved on with eyes averted, making genuine eye contact and a proper reply the exception rather than the rule — a stark contrast Rosenhan drew with how a passer-by on a university campus would answer the same question.
The pseudopatients' own accounts of what the experience did to them are a striking and often-overlooked part of the study. Several reported that, despite knowing full well they were sane and there voluntarily, they found the ward environment genuinely disorienting and demoralising — the boredom was profound, the loss of control corrosive, and the sense of being unseen wore at them within days. This is itself a finding: if a few days of the institution could unsettle healthy, forewarned volunteers who could leave the situation intellectually at any time, one can begin to imagine its effect on genuinely distressed patients with no such defences and no certain exit. The setting did not merely fail to help; it appeared actively to induce something of the passivity and hopelessness it was supposed to be treating.
Powerlessness pervaded the experience. Patients could be given medication (which pseudopatients, tellingly, disposed of, as did many real patients) with little explanation; their personal privacy was minimal, with living quarters, toileting and case notes open to staff; and staff would sometimes discuss patients in front of them as though they could not understand. Rosenhan documented an incident of a nurse adjusting her clothing in a day-room full of male patients — not out of cruelty, but because the patients simply did not register to her as people before whom one would not do so. The institution, he argued, systematically stripped away the ordinary marks of personhood.
The follow-up produced a mirror-image result. Over the three months, the hospital admitted many genuine patients. Of these, staff judged a substantial number to be suspected pseudopatients — roughly one in five patients was flagged as a likely fake by at least one staff member, and a proportion were doubted by a psychiatrist. Yet Rosenhan had sent none. Having primed staff to expect impostors, he had induced them to misclassify the genuinely ill as sane, just as readily as, in Study 1, they had misclassified the sane as ill.
| Finding | Study 1 (pseudopatients sent) | Study 2 (none sent) |
|---|---|---|
| The error made | Sane people diagnosed as insane | Genuinely ill people suspected of being sane |
| Detection by staff | None detected the sane | Many genuine patients wrongly flagged as fakes |
| What it shows | Diagnosis lacks validity toward "insane" | Diagnosis lacks validity toward "sane" |
Rosenhan drew a set of interlocking conclusions, captured in his famous line that "we cannot distinguish the sane from the insane in psychiatric hospitals."
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.