Cognitive Reflection and Decision Making
Introduces the Cognitive Reflection Test as a compact way to study whether people override tempting first answers.
Introduces the Cognitive Reflection Test as a compact way to study whether people override tempting first answers.
Frederick's paper made cognitive reflection easy to demonstrate. The original three-item Cognitive Reflection Test uses questions that invite an appealing first response. Solving them requires more than calculation: the learner must notice that the first answer should be checked before it is reported.
The paper presents the test as a compact measure of a person's ability or disposition to resist reporting the response that first comes to mind. It then compares test performance with choices involving time and risk, two areas that feature prominently in theories of decision-making.
Scores were associated with patterns in those choices and with other measures of cognitive ability or style. The result is useful because it separates possessing knowledge from initiating a check. A person may know every operation required by a problem and still answer incorrectly if the intuitive response ends the reasoning process too early.
Use reflection as a specific action rather than a vague instruction to “think harder.” Record the first answer, identify the assumption that makes it seem obvious, test the answer against every constraint, and only then commit. For numerical questions, estimate the direction and scale before calculating. For verbal claims, restate what would have to be true for the first interpretation to hold.
In Nuxflo, this supports confidence prompts, first-answer checkpoints, and explanations that show exactly why the tempting response fails. The goal is not to reward hesitation for its own sake; it is to build a reliable trigger for checking high-risk intuitions.
The CRT is narrow and familiarity with its famous items can change what a score means. Correlations with decision patterns do not make it a complete test of intelligence, wisdom, or everyday rationality. A short instrument can illustrate reflective override without measuring every form of careful reasoning.
Primary source: Journal of Economic Perspectives DOI.