Severe issues and solutions for the response inhibition data from the ABCD study and beyond
The stop-signal paradigm, a primary paradigm for measuring response inhibition, rests on the theoretical foundations of race models, which assumes independence between the overt go process and the inhibitory stop process. We show severe violations of independence across various implementations of the stop-signal task (Bissett, Jones, et al., 2021 Science Advances), but especially in the Adolescent Brain Cognitive Development [ABCD] dataset which encourages violations by embedding several experimental design issues (Bissett, Hagen, et al., 2021 eLife). We show how these violations can contaminate results and change existing conclusions in the literature. Finally, we propose solutions to address these violations, focusing on preliminary computational models to move towards a new theoretical framework for response inhibition.