ABSTRACT

Cognitive ability is neither static nor invariable. Just as human bodies display variability in size and shape, individuals manifest differences in cognitive abilities. The developmental dynamics and individual differences of cognitive ability have captured the interest of methodologists. The research discussed in this chapter applies a developmental method by McArdle and Nesselroade to item factor analysis of cognitive abilities. It focuses on the concept of Gf-Gc ability. The major interest is in developmental aspects of these two cognitive ability between adolescence and early adulthood age. The factor model is based on multivariate binary responses. The use of linear factor model techniques inevitably violates fundamental assumptions of linearity and multivariate normality tied with continuous variable factor models. Item response factor models require nonlinear structural equation modeling. McArdle and Nesselroade demonstrated how to model the longitudinal data, and the chapter extends such an approach to the longitudinal item data.