Then we replicated several prior associations between model-based planning and age, gender, education and specific clinical associations with compulsivity - adding more evidence to the validity of this gamified task 💎💎💎

.. but what can we do MORE with 📱's? In a series of A/B style manipulations, we found we could boost the signal between model-based planning and compulsivity when transition ratios were less deterministic & for certain reward probability drifts ⏫