This is a commentary on essays in this journal issue by M. E. Strauss (2001) and R. A. Knight and S. M. Silverstein (2001) on research methodology for studying cognitive deficits. Concentrating mostly on Knight and Silverstein's article, the authors review the psychometric issues in the matched-task design, analyze Knight and Silverstein's "process-oriented" objections to that design, and scrutinize their methods for studying cognitive deficits, examining 2 of their empirical studies as examples of those methods.