Individual Differences and the Normative-Descriptive Gap in Cognitive Biases
Differences in cognitive ability, task interpretation and thinking dispositions can help explain why judgments depart from normative models. The authors argue that computational limits and random errors do not account for all the systematic variation in their research.
THE IDEA
INDIVIDUAL DIFFERENCES AND THE NORMATIVE/DESCRIPTIVE GAP
We attempted to demonstrate that a consideration of individual differences in the heuristics and biases literature may have implications for debates about the cause of the gap between normative models and descriptive models of actual performance. Patterns of individual differences may alter our reflective equilibrium regarding the plausibility of the four alternative explanations for the normative/descriptive gap (performance errors, algorithmic limitations, wrong norm, and alternative construal).
Different outcomes were obtained across the wide range of tasks we have examined in our research program. Of course, all the tasks had some unreliable variance and thus some responses that deviated from the response considered normative could easily be considered as performance errors. However, not all deviations could be so explained. Several tasks (e.g., syllogistic reasoning with interfering content, four-card selection task) were characterized by heavy computational loads that made the normative response not prescriptive for some subjects – but these were usually few in number. Finally, a few tasks yielded patterns of covariance that served to raise doubts about the appropriateness of the normative models applied to them or the task construals assumed by the problem inventors (e.g., several noncausal baserate items, false consensus effect).
Although many normative/descriptive gaps could be reduced by these mechanisms, not all of the discrepancies could be explained by factors that do not bring human rationality into question. Algorithmic-level limitations were far from absolute. The magnitude of the associations with cognitive ability left much room for the possibility that the remaining reliable variance indicates that there are systematic irrationalities in intentional-level psychology. A component of our research program mentioned only briefly has produced data consistent with this possibility. Specifically, once capacity limitations are controlled, the remaining variation from normative responding is not random (which would have indicated that the residual variance consisted largely of performance errors). In several studies, we have shown that there is significant covariance among the scores from a variety of tasks in the heuristics and biases literature after they are residualized on measures of cognitive ability (Stanovich, 1999). The residual variance (after partialling cognitive ability) is also systematically associated with questionnaire responses that can be conceptualized as intentional-level styles relating to epistemic regulation (Sá et al., 1999; Stanovich & West, 1997, 1998c). These findings falsify models that attempt to explain the normative/descriptive gap entirely in terms of computational limitations and random performance errors. Instead, the findings support the notion that the normative/descriptive discrepancies that remain after computational limitations have been accounted for reflect a systematically suboptimal intentional-level psychology.
When to use it
In educational settings, design training that targets both cognitive skills and thinking dispositions to reduce biases.
In research, examine whether associations with thinking dispositions remain after statistical adjustment for measured cognitive ability, while considering alternative task interpretations.
In decision-making frameworks, incorporate reflective prompts to encourage more normative responses.
In consulting, assess both intelligence and thinking styles of clients to tailor interventions for better outcomes.
FROM THEORY TO EVERYDAY LIFE
In practice
Wason Selection Task
In the Wason selection task, participants choose cards to test a conditional rule. The source discusses the task as one in which computational demands can matter, while its broader research program also examines thinking dispositions and alternative interpretations. A wrong answer alone does not establish which explanation applies.
False Consensus Effect
The false-consensus effect concerns estimates of how widely others share one’s views. In this passage, its pattern of results raises questions about the normative model or the interpretation assumed by the researcher. It illustrates why a departure from the expected answer should not automatically be labeled irrationality.
A MOMENT TO REFLECT
Put it to work.
How do individual differences in cognitive ability affect performance on heuristics and biases tasks?
What evidence suggests that residual variance after controlling for cognitive ability is not random but related to thinking dispositions?
How might the normative/descriptive gap be explained by alternative construals of tasks rather than irrationality?
Often confused with
References
- The rationality debate and the normative/descriptive gap — Stanovich (1999)
- Individual differences in rationality: A dual-process account — Sá, W. C., West, R. F., & Stanovich, K. E. (1999)
- Individual differences in reasoning: Implications for the rationality debate — Stanovich, K. E., & West, R. F. (1997)
- Individual differences in rational thought — Stanovich, K. E., & West, R. F. (1998)
CONTINUE THE CONVERSATION
Source texts
- Heuristics and Biases The Psychology of Intuitive Judgment (Thomas Gilovich, Dale Griffin, Daniel Kahneman)