Check out our new paper, where we introduce g-distance to evaluate psychological models. Assessing not only how well models produce observed behaviours, but also how well they reject unobserved ones. Check it out at Psychological Review or read it here!