Why the Rosenhan study failed critics of psychiatry

The Rosenhan Pseudopatient Study, Its Unraveling, and Why the Case Against Coercion Never Needed It

The Rosenhan study is one of the most famous experiments ever published about psychiatry, and it now appears to be one of the most famous cases of research misconduct. This article is for learning and research. It uses the story to show how a weak study can damage the very cause it was meant to support, and why critics who see distress as problems in living rather than disease are on firmer ground when they argue from logic and human rights than from a single dramatic experiment. Getting that lesson right helps people solve real problems in living without depending on claims that can collapse.

This page is educational and historical. It is not medical, psychiatric, or legal advice.

What the 1973 paper claimed

In January 1973 the journal Science published a paper by psychologist David Rosenhan.[1] According to the paper, eight healthy volunteers, called pseudopatients, gained secret admission to 12 hospitals in five states on the East and West coasts. Each one reported a single symptom at intake: an unfamiliar voice that seemed to say "empty," "hollow," and "thud." Once admitted, each was told to stop reporting the voice and behave normally.

The paper reported that no staff member detected any of them. Eleven of the 12 admissions produced a diagnosis of schizophrenia, and the twelfth produced manic-depressive psychosis. Each pseudopatient left with a label of schizophrenia "in remission." Stays ranged from 7 to 52 days, with an average of 19 days. Other patients were said to be far more perceptive: during the first three stays, 35 of 118 patients on the admissions ward voiced suspicion that a pseudopatient was a journalist or researcher.[1]

A second part of the paper described a follow-up. A hospital was told that one or more pseudopatients would try to gain admission over the coming months. Staff rated 193 incoming patients, and 41 were judged with high confidence to be impostors by at least one staff member. Rosenhan wrote that no pseudopatient from his group had actually been sent.[1]

The conclusion was sweeping. Diagnosis, the paper argued, told observers more about the setting than about the person, and the hospital itself turned ordinary behavior into evidence of illness.

How the study came apart

Critics raised method problems early. Psychiatrist Robert Spitzer published a long critique in 1975 arguing that the paper misused logic and data.[2] In 1976 a ninth pseudopatient, Harry Lando, published a supplemental report describing his own stay, which the original paper had left out.[3]

The larger break came decades later, when journalist Susannah Cahalan searched Rosenhan's private files for her 2019 book The Great Pretender. In her own account she reports several findings:[4][5]

  • Rosenhan's own hospital record did not match the single-symptom script. According to the record, he described a much more elaborate story, including covering his ears with copper pots, sensitivity to radio signals, and thoughts of suicide.
  • Lando had a positive experience. He described a caring staff and a healing ward. Rosenhan dropped him in a footnote for falsifying his personal history, although other pseudopatients had also altered their histories.
  • An earlier draft with nine pseudopatients listed the same average stay and the same counts as the published version with eight. Not one number changed after Lando was removed.
  • Cahalan could confirm only a few of the pseudopatients, which raised the question of whether some existed at all.

A review in The Lancet Psychiatry summarized her work as showing serious discrepancies between Rosenhan's notes and the published paper.[6] In 2023 historian Andrew Scull, working from records Cahalan shared with him, described the study in History of Psychiatry as "a spectacularly successful case of scientific fraud."[7]

The irony: a critique that strengthened psychiatry

The most surprising part of the story is its effect. Scull writes that the paper played a crucial role in persuading the American Psychiatric Association to revise its diagnostic manual, and that the third edition (DSM-III) launched a revolution in American psychiatry whose effects continue today.[7] Rosenhan himself ended his paper by hoping for community services that avoid psychiatric labels.[1] Instead, the profession answered the embarrassment with checklist diagnosis. Reliability became the goal, and more precise labels became the solution to a problem that the paper had framed as being about labels in the first place.

In other words, a study meant to expose the power of diagnosis helped make diagnosis more systematic, more confident, and more central. Scull also notes that the paper is still cited hundreds of times a year as evidence of the failings of psychiatry.[7] Each of those citations is now a liability for anyone making that argument.

A Szaszian reading

Thomas Szasz published "The myth of mental illness" in 1960, thirteen years before Rosenhan. His argument did not rest on whether doctors could spot fake patients. It rested on what the word "illness" means. Szasz argued that psychiatry deals mainly with problems in living, conflicts over needs, values, and relationships, and that calling these problems diseases gives an impersonal "thing" as an explanation for what are really human and moral struggles.[8]

That argument is conceptual. It cannot be refuted by a better checklist, and it cannot be saved by a dramatic experiment. Seen this way, the collapse of the Rosenhan study changes less than it first seems:

  • It does not show that psychiatric diagnosis is valid. Fraud in a study that attacked diagnosis is not evidence for diagnosis.
  • It does not touch the central human rights concern, which is that people can be detained and treated against their will on the basis of a label. That concern is about power and consent, not about detection rates. See Psychiatric coercion and human rights.
  • It does show that critics need standards at least as high as those they demand of psychiatry. A movement that leans on a weak study hands its opponents an easy win.

Lando's account adds a further point that is easy to miss. Some people report good experiences inside institutions. A serious critique can accept that and still argue that consent, not the quality of a particular ward, is the moral line.

Lessons for learning and research

The Rosenhan story is useful well beyond psychiatry, because it shows how a famous result can travel for decades without being checked.

  • A memorable story beats a careful one. Three words, "empty, hollow, thud," made the study easy to teach and easy to repeat. Memorability is not evidence.
  • Look for the excluded data. Lando's positive stay sat in a footnote. Asking who was left out, and why, is one of the most productive questions in any review.
  • Unchanged numbers are a warning. When removing a participant changes nothing in the averages, something is wrong with the averages.
  • Primary records matter. The problems surfaced only when someone compared hospital records with the published text.
  • Replication markets and open data help. Approaches discussed in Prediction markets for replication aim to put a price on doubt before a result spreads.

Limits and open questions

  • Much of what is now known comes from one investigator's archive work, summarized by others. Not every record is public, and some of Cahalan's conclusions, such as whether certain pseudopatients existed, remain inferences.
  • The paper's account of depersonalization in hospitals was not the part most clearly disproven. Other evidence on coercion and conditions in psychiatric care should be judged on its own merits, not dismissed because this study failed.
  • The claim that the paper drove the DSM-III revision is a historical judgment made by Scull, and a single paper is rarely the only cause of a major reform.
  • Explain how a study meant to weaken psychiatric labeling may have helped make diagnosis more central to American psychiatry.
  • Compare the strength of an argument that rests on a field experiment with one that rests on the meaning of the word illness, as Szasz argued.
  • Describe what Harry Lando's positive hospital stay adds to the debate, and whether a good experience inside an institution changes the case against coercion.
  • Identify warning signs in a famous study that a careful reader could have noticed before the original records were examined.
  • Discuss what standards of evidence critics of psychiatry should hold themselves to if they want their arguments to last.
  • Propose a way that open data or replication markets could have caught the problems in the Rosenhan study decades earlier.
  • Choose another famous result in psychology that is widely taught and outline how you would check it against primary records.

See also

References