Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster A81 in Poster Session A: Tuesday, August 12, 1:30 – 4:30 pm, de Brug & E‑Hall

Computational Modeling of Choice Frequency in Habitual Behavior: A Pre-Registered fMRI Study

Hugo Fluhr¹, Viktor Timokhov¹, Philippe Tobler, Stephan Nebe; ¹University of Zurich

Presenter: Hugo Fluhr

Habits are integral to decision-making but can contribute to maladaptive behavior when they override goal-directed control. Previous research showed that past choice frequency influences current choice independently of reward history. However, many prior attempts to experimentally induce habitual behavior have failed to replicate (e.g., Gera et al., 2023), highlighting the need for robust behavioral paradigms. In this pre-registered fMRI study (N = 71), we aimed to replicate and extend previous behavioral findings to different behavioral situations. Moreover we aimed to investigate the neural mechanisms underlying frequency-based habitual behavior. Participants completed a modified version of the Reward Pairs task, where reward level and choice frequency were manipulated orthogonally. Behavioral data were best explained by a model combining reinforcement learning (RL) and a choice kernel (CK), confirming that both reward and choice history shape current choices. Our pre-registered univariate fMRI analyses are ongoing. So far, they revealed no significant neural correlates of RL or CK values in hypothesized regions of interest. While the behavioral findings reinforce the relevance of frequency-based habits, the neural data point to challenges in detecting their neural substrates using univariate BOLD analyses. Future work will examine whether these effects manifest through distributed neural patterns or in the functional connectivity between candidate brain regions.

Topic Area: Reward, Value & Social Decision Making

Extended Abstract: Full Text PDF