Large prospective losses lead to sub-optimal sensorimotor decisions in humans

The rationality of human behavior has been a major problem in philosophy for centuries. The pioneering work of Kahneman and Tversky provides strong evidence that people are not rational. Recent work in psychophysics argues that incentivized sensorimotor decisions (such as deciding where to reach to get a reward) maximizes expected gain, suggesting that it may be impervious to cognitive biases and heuristics. We rigorously tested this hypothesis using multiple experiments and multiple computational models. We obtained strong evidence that people deviated from the objectively rational strategy when potential losses were large. They instead appeared to follow a strategy in which they simplify the decision problem and satisfice rather than optimize. This work is consistent with the framework known as bounded rationality, according to which people behave rationally given their computational limitations.