Applying Log-Linear Models to the Study of Career Development and Transition of Individuals With Disabilities

The analysis of categorical data has relied heavily on the traditional chi-square test. Although useful, chi-square is limited to testing single pairs of variables. Over the past several decades a powerful extension to chi-square analysis, able to analyze associations among several variables simultaneously, has been developed. This multivariate extension, known as log-linear analysis, offers opportunities for investigators to analyze and interpret categorical data with greater precision than previously possible. We provide an overview of the major steps involved in log-linear analysis and seek to promote its use when categorical data are lcollected from studies on the career development and transition of individuals with disabilities from school to adult life.