Publication

A Bayesian hierarchical model of trial-to-trial fluctuations in decision criterion

Every day, we make numerous decisions, from choosing what to eat to how we interpret the world around us. Traditionally, researchers have assumed that a key part of our decision-making process, namely the decision criterion, stays stable over time. However, increasing evidence suggests that instead of being stable, the decision criterion can fluctuate from moment to moment.

Author: Anne E. Urai, Robin Vloeberghs, Kobe Desender, Scott W. Linderman
Date: 29 July 2025
Links: Public Library of Science (PLOS)

Fig 1. The consequences of fluctuations in decision criterion.

In this paper, authors at KU Leuven, Leiden University and Stanford University introduce the Hierarchical Model for Fluctuations in Criterion (hMFC), a new computational model that can accurately estimate these fluctuations, even with limited data. Capturing these moment-to-moment changes is critical: ignoring them can bias estimates of perceptual sensitivity and the influence of past choices on current decisions. By estimating criterion fluctuations, hMFC can correct these biases. With hMFC, we offer researchers a powerful tool for uncovering the dynamic nature of human decisions.

Read the full publication through the Public Library of Science.