Simple one-reason heuristics can outperform complex regression

Correct: B — Match or out-predict more complex strategies in real-world environments while using less information

Gigerenzer & Goldstein (1996) compared 'Take the Best' — a one-reason decision heuristic that uses cues in order of validity and stops at the first discriminating cue — against weighted-additive models (including multiple regression) on real-world prediction tasks (e.g., German city populations). In out-of-sample prediction, the simple heuristic matched or beat the more information-greedy strategies. The work helped reframe cognitive heuristics from 'biases that hurt' (the heuristics-and-biases tradition) to 'ecologically rational adaptations' that exploit redundancy and noise in the environment. The framework grounds modern adaptive-toolbox theories of decision-making.