Validating an experimental paradigm for inducing and measuring habits

Habits are difficult to induce and measure in humans using laboratory-based tasks. Experimental paradigms that reliably demonstrate overtraining-induced inflexibility, a hallmark of genuine habits (as opposed to e.g. weak goal-directed control), mostly fail to do so in humans \citep{tricomi2009specific, pool2022determining, gera2023characterizing, dewit2013resolution, dewit2018shifting}. This raises the possibility that in humans, these experiments target cognitive processes besides habit formation and habitual control. Indeed, multiple purported task-based behavioral indices of habits appear not to relate to each other, nor to questionnaire-based measures of habits \citep{friedel2014devaluation, gillan2015model, sjoerds2016slips, linnebank2018investigating, nebe2024characterizing}. We recently developed an experimental paradigm that reliably exhibits increased behavioral inflexibility in overtrained participants within a single $\sim$1-hour session, without explicitly limiting the amount of goal-directed (GD) control participants can exert \citep{oh2025naturally}. Thus, we believe that in overtrained participants, our task measures the natural expression of gradually-formed, genuine habits. The aim of this project is to show that the overtraining-induced inflexibility observed in our behavioral paradigm is the result of the same gradual habitization process that drives real-life habit formation. Successfully linking behavioral indices of habit extracted from our task to external measures that should relate to real-life habits would validate our paradigm's potential as a useful tool for studying habits in the laboratory. See https://osf.io/bsp6f/files/576yw for a PDF version of this preregistration.