Answers are grounded in this paper and cite their sources.
Shows interactions between choice serial effects and confidence ratings in diverse tasks, suggesting shared mechanisms.
Perception and memory at timepoint N are biased by the timepoint N-1 decision, a phenomenon known as serial dependence. A similar dependence has been found for confidence ratings, a phenomenon known as “confidence leak.” These two types of serial effects have mostly been assumed to be independent and have typically been attributed to distinct explanatory mechanisms. Here, using data from 26 experiments (total N = 1876), we show that choice and confidence serial effects are strongly associated and even modulate each other. We first confirm that choice and confidence serial effects are robustly present in both perception (16 experiments) and memory (10 experiments) tasks. Critically, we find that choice serial effects and confidence leak interact with each other in all 26 experiments. Specifically, switching the perceptual or memory choice substantially weakens confidence leak. What is more, switching the confidence rating not only attenuates but in fact reverses the choice serial effect. Further, we show that this interaction between serial effects occurs at the level of sensory evidence and that causal manipulations of one type of serial effect affect the other. Finally, we show that subjects who exhibit stronger choice serial effects also exhibit stronger confidence leak. These results demonstrate that serial effects in choice and confidence are strongly related and suggest the possibility of common mechanisms underlying different serial effects.
No takes yet. Share an insight, caveat, or question.
Bocheva et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: