← The Casebook

An examined document · read the margins

Congruence Bias

The tendency to evaluate hypotheses by testing cases expected to yield a positive answer if the favored hypothesis is true, while neglecting tests of alternative hypotheses that could reveal it to be false.

Filed · congruence-biasAnnotated in the gutter

Jonathan Baron, Jane Beattie, and John Hershey formally identified congruence bias in 1988 through six experiments on diagnostic question evaluation. Subjects systematically overvalued questions with high probability of yielding 'yes' under their most likely hypothesis, even when such questions were normatively uninformative. This bias was reduced when alternative hypotheses or probabilities of negative results were made explicit. The concept extends Wason's (1960) earlier 2-4-6 task findings, where subjects tested only confirming instances of hypothesized rules.

Congruence bias arises from a failure to consider arguments against an initial possibility. When evaluating a question, people attend to the probability of a positive answer under the favored hypothesis but fail to consider whether the same question distinguishes between alternative hypotheses or whether a negative answer would be equally informative. Baron (1985) proposed that this reflects a more general tendency to end the search for evidence too early. Positive information is cognitively more accessible than negative information, making congruent tests the default.

Klayman and Ha (1987) argue that what appears to be congruence bias is better understood as a generally adaptive positive test strategy that becomes maladaptive only in specific task structures (when the hypothesized rule is embedded within the correct rule). Evans (1989) suggests the bias reflects difficulty thinking about negative information rather than a desire to confirm.

Further annotations

Wason Four-Card Selection Task. Most participants selected the confirming cards (vowel and even number) rather than the logically necessary combination (vowel and odd number). Only about 10% selected the correct cards. This pattern of selecting confirmatory over falsifying evidence has been replicated extensively.

Wason, 1966 · not reported

case · Medical Diagnostic Testing and Unnecessary Procedures (1988)

Baron et al. (1988) cited evidence that the information heuristic (a related bias) contributes to medical professionals ordering tests that are expensive, dangerous, and unnecessary. Allman et al. (1985) and Elstein et al. (1978) documented cases where physicians ordered tests that could not change the treatment decision, consistent with congruence bias in diagnostic reasoning.

case · Keren and Fischhoff Diagnostic Reasoning Studies (1988)

Keren and Fischhoff (1988) found that in medical diagnostic scenarios, subjects selected questions likely to yield positive results under their favored hypothesis but undervalued questions testing alternative diagnoses, leading to inefficient verification processes. This pattern is directly analogous to congruence bias in clinical settings.

Catch it in the act

Watch for situations where you are asking only questions likely to yield 'yes' if your preferred explanation is true. In medical contexts, notice when all ordered tests would only confirm the suspected diagnosis without ruling out alternatives. In problem-solving, check whether you are testing only cases that fit your theory.

Document of record from Jonathan Baron, Jane Beattie, John C. Hershey, 1988 — Heuristics and biases in diagnostic reasoning: II. Congruence, information, and certainty.

Cross-references in the margin

File your own case

Open the same case on your own draft.

Paste a memo, a research draft, or a strategy argument. It is scored against all 175 cards, and the strongest two or three risks come back with the evidence quoted and one practical next check.

Open a case on your draft →