Mental chronometry and individual differences in cognitive processes: Common pitfalls and their solutions

Abstract

Three general problem areas are illustrated for the use of Chronometrie data and modeling techniques in the domain of basic mathematics performance. The focus of the article is on appropriate and inappropriate uses of latency data to determine which strategies govern an individual's performance on a set of specific mathematics basic facts problems. The three general problem areas in applying Chronometric methods are (1) data aggregation and model fitting, (2) covariance and relationships among models, and (3) strategy mixture. All three problem areas render it difficult to analyze latency data from the perspective that there is a single best model that characterizes an individual's performance. Solutions to this dilemma are provided in the form of methods of data collection and/or analysis procedures that explicitly examine strategy mixture. Consideration is given to the utility of analyzing multiple strategies in addressing issues of individual differences in cognitive skills.