The Number of Response Options as a Determinant of Decision-Making Quality

Insights into decision-making are often made via the use of binary response tasks. We ask whether such behavioural signatures are generalizable to other cases of non-binary, and perhaps more realistic, decision-making environments. In Experiment 1 (n = 95), participants played four zero-sum games with different numbers of response options (2, 3, 5, 6) against an unexploitable computerized opponent. In Experiment 2 (n = 69), participants played the same games against an exploitable opponent. The data from Experiment 1 – 2 confirmed that more response options promoted random explorative behaviour, the binary response game promoted win-stay behaviour while the non-binary response games promoted lose-shift behaviour, and, that post-win slowing were independent of response availability. In Experiment 3 (n = 72), participants played the same zero-sum game with different numbers of response options (2, 4, 6) for both players, enabling us to identify two types of shifts behaviour: rule-relevant shift (indicative of strategic change) and rule-irrelevant shift (indicative of random exploration). The data showed that rule-relevant shift increased in function of response options only following wins but not losses, demonstrating cognitive inflexibility following negative outcomes. In contrast, rule-irrelevant shift was independent of outcome valence, suggestive of information processing noise, and was promoted by more response options. These results emphasize response availability as an important determinant of decision-making behaviour and highlights the limitations of binary-response tasks. Future work should continue to investigate how response availability modulates neurophysiological responses during such explore-exploit dilemmas.