Underlying Construct Analysis (UCA): Where Individual Systems End and Human Systems Begin

The association of antecedents to performance outcomes, including individual behaviors, interfaces and the environments in which engagements are performed and sustained as a behavioral system is certainly implied, but has not been a primary focus of traditional data mining techniques. The current paper introduces the concept of Underlying Construct Analysis (UCA) using techniques associated with confirmatory factor analysis, and demonstrates their utility for theory development. We provide step-by-step instructions for data mining beyond observable variables, to include larger units of behavior frameworks and demonstrate using a case example, unique variance associated with two distinct populations in a large data set (Survey of Parents and Children, 1990), populations that were not defined prior to the UCA. We lastly emphasize the potential global application of our proposed procedures for individuals with only moderate levels of understanding of statistics, and across multiple domains of exploration.