Reading focus
Syllabus: How Do We Reason and Solve Problems?
Keep the choice of a test separate from whether its outcome confirms or contradicts a hypothesis. Compare environments in which the same strategy succeeds or fails.
Why you're reading it. Replace a broad bias label with an account of information search and its conditions.
A question to consider. After three successful deliveries, someone hypothesizes that a route is always reliable. What rival explanations should determine the next delivery conditions to investigate?
Why this reading
Testing a case that a hypothesis predicts should succeed is not necessarily an attempt to avoid discovering an error. This analysis separates the choice of a test from its possible outcomes and asks when a familiar testing strategy actually supplies useful information.
Before you read
A positive test examines a case expected to possess the property of interest. Confirmation and disconfirmation describe what the evidence does to a hypothesis. A positive test can disconfirm it. In a deterministic task, a rule has no exceptions; in a probabilistic setting, a result can change confidence without decisively settling the issue. The set diagrams describe how a proposed rule overlaps the true set of cases.
Read for
- Track the distinction between a testing strategy and a preference for favorable outcomes. Find a positive test that could expose a false hypothesis.
- Compare hypotheses that are narrower than, broader than, or only partly overlap the true rule. Explain why the same strategy has different consequences.
- Follow how the information supplied by a test depends on the environment, uncertainty, and available alternatives. Evaluate the authors' claim about a useful default strategy separately from claims that it is always best.
Consider afterwards
- You think a device accepts every even-numbered code. Testing 8 succeeds. Propose two rival rules still consistent with that result and choose a next test that distinguishes them. Would a positive test necessarily be uninformative?
Editorial context
This is a theoretical reanalysis and synthesis, not a demonstration that confirmation bias never occurs. Its contribution is to specify which behavior needs explaining and the conditions under which it helps or hurts. Deliberate avoidance of contrary evidence, biased interpretation of a result, and selection of a positive test should not be collapsed into one mechanism. The formal treatment makes this a demanding but useful conceptual reading.