Why this reading
Spitzer challenges the inferential steps in Rosenhan's report, including what admission after reported hallucinations and discharge with a diagnosis in remission can show. This optional companion reading teaches how to examine a study's logic without merely accepting or rejecting its conclusion. Read the two papers together.
Before you read
Identify the claims about admission, labeling, and recognition of improvement that Spitzer attributes to Rosenhan, then distinguish the original claims from Spitzer's interpretation of them. Comparing the two papers helps assess whether the critique addresses the report's evidence and conclusions fairly.
Notice Spitzer's position: he was at this time leading the development of DSM-III. He is an interested party, and part of the exercise is deciding whether that affects the strength of the argument.
Read for
- The distinction Spitzer draws between the validity of a diagnostic category and the behaviour of the people who apply it.
- His treatment of the "in remission" discharge. Why does he regard it as the most telling detail in the study?
- The argument that Rosenhan's design tests the detection of deliberate deception, not the validity of diagnosis, and what a study of validity would need instead.
- The rhetoric. The title accuses Rosenhan of pseudoscience; decide whether the argument earns the accusation.
Consider afterwards
- Which of Rosenhan's claims survive Spitzer's critique, if any? Is the labelling claim damaged by the same reasoning as the detection claim?
- Spitzer's remedy for unreliable diagnosis was explicit criteria. Does his critique of Rosenhan depend on that remedy being right?
Editorial context
The critique argues that the report's conclusions do not follow from its design. That methodological case should be evaluated on its merits, rather than presented as an uncontested verdict on every institutional or clinical issue raised by Rosenhan. It also does not independently verify the original observations. Compare the arguments about diagnostic error, labels, and recognition of improvement before deciding which claims each paper actually addresses.