What's the goal? Reconsidering the importance of higher-order cognition in internalizing psychopathology

To date, the field of computational psychiatry has largely focused on testing for dysfunction within lower-level cognitive processes in psychopathology, such as biases related to attention, reinforcement learning and memory. While this work has yielded important insights, there is clear evidence from empirically validated psychotherapeutic interventions to suggest that these lower-level cognitive processes are often intact. Here, we argue that etiological and maintaining factors for internalizing disorders may be more appropriately situated in higher level cognitive processes such as goal-selection. We further suggest that future research should incorporate behavioral tasks and models that allow for participant-defined goals to reveal higher-level mechanistic targets. We believe that these higher-level deficits in goal-selection, possibly in tandem with low-level impairments, may augment ecological validity and clinical relevance.