I-states-as-objects-analysis (ISOA): Extensions of an approach to studying short-term developmental processes by analyzing typical patterns

Abstract

I-states-as-objects-analysis (ISOA) is a person-oriented methodology for studying short-term developmental stability and change in patterns of variable values. ISOA is based on longitudinal data with the same set of variables measured at all measurement occasions. A key concept is the i-state, defined as a person’s pattern of variable values at a specific time point. All i-states are first subjected to a classification analysis that results in a time-invariant classification characterized by a number of typical i-states. Each person is then characterized at each time point by the typical i-state he/she belongs to. Then the person’s sequences of typical i-states are analyzed with regard to structural and individual stability and change. Extensions of ISOA are presented where: (1) some methods for checking the assumption of a time-invariant classification are indicated; (2) information about the degree of dissimilarity between typical i-states is used as an aid in interpreting the findings; and (3) attention is given to closed paths—that is, typical i-state sequences that do not occur at all. To demonstrate the methods, an empirical example is given that concerns the development of children’s achievement-intelligence patterns between ages 10 and 13.