Individual differences are a fundamental component of psychology, but these differences are often treated as “noise” or “errors” in variable-oriented statistical analyses. Currently, there is a small but emerging body of research using the person-oriented approach. In this paper a brief theoretical and methodological overview of the person-oriented approach is given. A person-oriented approach is often preferable where the main theoretical and analytical unit is a pattern of operating factors, rather than individual variables. In order to illustrate the relevance of this approach to research in educational psychology several representative statistical methods are outlined, two of which employ a person-oriented approach (latent class analysis/ latent profile analysis, configural frequency analysis/ prediction configural frequency analysis) and one that combines person and variable-oriented approaches. Examples of data analyses are used to demonstrate that variable and person-oriented approaches provide the researcher with different information that can be complementary.
Key words: configural frequency analysis, educational psychology, individual differences, latent class analysis, person-oriented approach.
Local Associations in Latent Class Analysis: Using Configural Frequency Analysis for Model Evaluation
