Pseudo-Optimal Evidence Accumulation as a Foundation for Multi-Choice Irrationalities
Kiyofumi Miyoshi, Yosuke Sakamoto · 2024
Everyday decision-making often involves choosing between multiple alternatives, and establishing unified theories for this process would benefit broad disciplines of behavioral science. A key challenge lies in explaining behavioral irrationalities that specifically arise in multi-alternative decisions. This study, based on pre-registered procedures, investigated such non-normative behaviors in a three-choice dot numerosity discrimination task. We systematically manipulated the value of the lowest-numerosity alternative (the dud stimulus) and quantified its non-normative effects on choice, response time, and confidence. We then explained observed irrational behaviors with a sequential evidence accumulation model featuring a “max vs. next” decision algorithm. This pseudo-optimal decision-making rule naturally captured a well-known non-normative choice pattern: the violation of the independence of irrelevant alternatives. Our model also explained a multi-choice metacognitive bias, known as the dud-alternative confidence boost, by considering evidence accumulation dynamics right after stimulus presentation. By uncovering the computational processes behind these non-normative multi-choice behaviors, this study represents a significant advancement in understanding the core principles of decision-making.