Potential Contributions of Clinical Mathematical Psychology to Robust Modeling in Cognitive Science

Potential contributions of clinical mathematical psychology to robust modeling in cognitive science are described. Potential contributions include model generalization testing, through evaluating model performance with extreme individual differences provided by clinical samples. Solution-oriented model support, in the form of end-use vindication, is available by exploiting measurement models for clinical assessment of symptom-significant cognitive functioning and monitoring cognitive aspects of treatment regimens, notably CNS-directed pharmacotherapy. Provision can be made for formal, transparent anchoring of cognition in clinical cognitive neuroimaging to counter “reverse logic” (circularity) in clinical neuroimaging research. Analytical modeling also can provide a quantitative nexus for integrating levels of neuroimaging (e.g., functional magnetic resonance imaging and functional magnetic resonance spectroscopy). Mixture models, treating model parameters as being randomly distributed across participants, defensibly can extend model support to occurrences of “data over-dispersion”. Model support, furthermore, is available according to theoretical significance of parameter mixing distribution hyper-parameters. Parameter mixing distributions furthermore can serve as stabilizing Bayesian priors, to help address clinically imposed small N model testing. Although emanating from clinical mathematical psychology, appropriation of several recommended practices, to quantitative cognitive modeling generally, is deemed advisable.