Case file · AP Psychology · AAQ practice

Article Analysis Question · Unit preview: Clinical psychology

On Being Sane in Insane Places

Eight people with no history of mental illness got themselves admitted to psychiatric hospitals. Once inside, they acted completely normal. The question Rosenhan asked: would anyone notice?

PatientPseudopatient  XXXXXXX 
FacilityHospital 1 of 12
ComplaintHears voices: “empty,” “hollow,” “thud.”
Admitting DxSchizophrenia
Ward notes“Patient engages in writing behavior.”
Discharge DxSchizophrenia, in remission

Intake form

Check in before rounds

Training wheels

Ward rounds completed

Your answers save in this browser as you go. When you finish, download your lab as a PDF at the bottom of the page and upload it to the Canvas assignment.

Before you open the file

Make a prediction

Eight healthy adults call psychiatric hospitals and report one symptom: hearing a voice say “empty,” “hollow,” and “thud.” If they get admitted, they stop faking right away and act completely normal. Commit to a guess before you read on.

Case summary

What Rosenhan did, and why it shook psychiatry

In the early 1970s, psychologist David Rosenhan wanted to test a simple-sounding question: can mental health professionals reliably tell the sane from the insane? If diagnosis is based on the patient, it should not matter where the patient is. If it is shaped by the setting and the label, healthy people might be “seen” as ill just because they are inside a psychiatric hospital.

Eight mentally healthy adults (including Rosenhan) called psychiatric hospitals and reported a single symptom: hearing an unfamiliar voice say “empty,” “hollow,” and “thud.” Every one of them was admitted. Right after admission, they stopped reporting any symptoms and behaved normally, cooperating with staff and taking notes on what they saw.

No staff member ever detected a pseudopatient. Most were diagnosed with schizophrenia and were released with the label “schizophrenia in remission,” meaning the label stayed with them even when the symptoms had “gone away.” Ordinary behavior, like note-taking, was recorded as a symptom. Ironically, many of the real patients suspected the pseudopatients were faking.

8 of 8pseudopatients admitted; none detected by staff
12hospitals in 5 U.S. states, East and West coasts
19 daysaverage stay (range 7–52 days)
35 of 118real patients voiced suspicion during the first three stays

Three ideas to carry into the clinical unit

  • The nuances of diagnosis. Diagnosis relies on what a patient reports and how clinicians interpret it. Context can tilt that interpretation.
  • Operational definitions. What exactly counts as “sane”? If a variable is not defined in observable terms, two observers can see the same behavior and reach opposite conclusions.
  • The power and damage of a label. Once a person is labeled, everything they do can be read through that label. Rosenhan called this the “stickiness” of psychodiagnostic labels.

Terms you’ll see again in the clinical unit

The chart vocabulary

Labeling

Attaching a diagnostic name to a person, which can change how others see them and how they see themselves.

In Rosenhan: the label stuck even at discharge, as “schizophrenia in remission.”

Confirmation bias

Noticing and interpreting information in ways that confirm what we already believe.

In Rosenhan: note-taking became “writing behavior,” a symptom.

Reliability of diagnosis

Whether different clinicians give the same diagnosis to the same person.

In Rosenhan: the same complaint led to schizophrenia 11 times and manic-depressive psychosis once.

Validity of diagnosis

Whether a diagnosis accurately reflects a real condition the person has.

In Rosenhan: healthy people were diagnosed with a serious disorder.

False positive / false negative

Calling something present when it isn’t (Type I) or missing something that is there (Type II).

In Rosenhan: healthy people called ill, then real patients called fakes in the follow-up.

DSM

The Diagnostic and Statistical Manual of Mental Disorders, the standard U.S. system for classifying disorders.

After Rosenhan: the DSM-III (1980) added specific, observable criteria to improve reliability.

Depersonalization

Being treated as an object or a case instead of as a person.

In Rosenhan: staff walked past patients’ questions most of the time.

Operational definition

Defining a variable by exactly how it will be observed or measured.

In Rosenhan: “responsiveness” was coded by staff reactions to a question. “Sane” never was.

Exhibit A · Source for the AAQ

The article, in AP exam format

On the exam, the article summary is organized into sections like these. Every answer in Parts A–F comes from evidence in this exhibit, so keep it handy.

Introduction

Rosenhan tested whether the characteristics that lead to psychiatric diagnoses reside in the patients themselves or in the environments and contexts in which observers find them. He hypothesized that if healthy people were admitted to psychiatric hospitals, staff would fail to recognize them as sane, because diagnoses depend more on the setting and the label than on the patient’s actual behavior.

Participants

Eight pseudopatients (three women and five men), including a psychology graduate student, three psychologists, a pediatrician, a psychiatrist, a painter, and a homemaker. None had a history of mental illness. The hospitals they entered (12 total) were in five states on the East and West coasts of the United States. They ranged from old to new, research-oriented to not, understaffed to well staffed; all but one were public (state, county, or university-funded) and one was private.

Method

Each pseudopatient called a hospital for an appointment and told the admitting staff that they heard an unfamiliar, same-sex voice saying “empty,” “hollow,” and “thud.” They gave a false name and occupation, but all other life history was reported truthfully. Immediately after admission, they stopped simulating any symptoms and behaved as they normally would. They were not told when they would be released; each had to get out on their own by convincing staff they were well.

While on the ward, pseudopatients openly took notes on ward life. To measure staff responsiveness, a pseudopatient approached a staff member with a brief, courteous question (for example, “Pardon me, Dr. X. Could you tell me when I am likely to be discharged?”) and recorded the staff member’s response as one of the following: moved on with head averted, made eye contact, or stopped and talked. The names of the hospitals and staff members were not disclosed in the report.

Results

All eight pseudopatients were admitted. Eleven of the twelve admissions were diagnosed as schizophrenia, and one (at the private hospital) as manic-depressive psychosis. No pseudopatient was detected by staff. Each was discharged with a diagnosis of schizophrenia “in remission.” Hospital stays ranged from 7 to 52 days, with an average of 19 days. During the first three hospitalizations, 35 of the 118 real patients on the wards voiced suspicion, saying things like “You’re not crazy. You’re a journalist, or a professor.” Pseudopatients were given a total of nearly 2,100 pills. Contact with psychiatrists averaged under 7 minutes per day. Nurses recorded normal note-taking as a symptom: “Patient engages in writing behavior.”

Staff responses to pseudopatients’ questions
ResponsePsychiatristsNurses & attendants
Moved on, head averted71%88%
Made eye contact23%10%
Stopped and talked≈6%≈2%

In a follow-up, staff at a teaching hospital who had heard about the results were told that one or more pseudopatients would try to be admitted over the next three months. Staff rated each new admission. Of 193 patients judged, 41 were judged with high confidence to be pseudopatients by at least one staff member, and 23 were suspected by at least one psychiatrist. In fact, no pseudopatients were sent.

Discussion

Rosenhan concluded that “we cannot distinguish the sane from the insane in psychiatric hospitals.” Once a person was labeled, staff interpreted normal behavior as part of the disorder, and patients experienced powerlessness and depersonalization. He argued for community mental health care and for diagnosing behaviors rather than attaching global labels to people.

Before you begin

How this practice works

On the real exam, there are no training wheels.

You will write every sentence of the AAQ yourself, with no menus and no word bank. The AAQ is worth 7 points and you’ll have about 25 minutes for it in the free-response section.

Training wheels on. For each part, read the charge (the actual task), then what it means and where to look. Build the sentence with the drop-downs, then press Check this part. Every check counts as an attempt, so think before you click. Your first-try results go in your PDF. Clear answers empties a part so you can retry it, but your check count and your Reader’s desk calls stay.

Once a part is cleared, the Reader’s desk opens: you score two sample student answers the way an AP reader would.

Training wheels off. For each part, read the charge and write your answer in the box, the same way you will on exam day. Then press Compare with model answer and score yourself honestly. The Reader’s desk opens after you compare.

Want the real conditions? Start the 25-minute exam timer.

Debrief · Group discussion

The power and damage of a label

The study was a bombshell, and it helped push psychiatry toward clearer, more operational diagnostic criteria (the DSM-III in 1980). It has also drawn serious criticism. Psychiatrist Robert Spitzer argued that a hospital is right to admit someone reporting hallucinations, and that the rare discharge label “in remission” shows staff did notice the absence of symptoms. In 2019, journalist Susannah Cahalan’s book The Great Pretender raised questions about whether Rosenhan reported his data accurately, and she could locate only some of the pseudopatients. Keep both sides in mind: a famous study is still a claim to be checked.

Discharge

Download your lab for Canvas

The PDF includes your name, your prediction, each part’s response, how many checks each part took, your own writing, your Reader’s desk scores, and your discussion answers.