Commentary: Building the developmental foundations of developmental computational psychiatry: reflections on Hauser et al. (2019).

There is a growing interest in applying the conceptual and analytical frameworks of computational psychiatry to developmental populations. This is motivated by appreciation that psychiatric illness needs to be understood from a neurodevelopmental perspective. The target article by Hauser and colleagues highlights progress in applying the computational psychiatry perspectives to identifying the developmental mechanisms of mental illness. We share the enthusiasm and optimism for this venture, while recognizing the substantial theoretical and pragmatic challenges associated with applying computational frameworks to developing populations. In this commentary, we highlight the ways that taking a developmental perspective in this arena stretches beyond merely identifying age differences in a computational parameter of interest. These include the need for experimental and computational frameworks to recognize that developmental changes can be quantitative or qualitative in nature, the need to consider developmental stage beyond age groupings or even numerical age, and the need for large quantities of data to model age-related changes in a reproducible manner. In doing so, we hope to stimulate progress in uncovering the mechanisms of psychiatric illness in a way that is developmentally informed.