In education and other social science research, researchers often encounter data with multilevel structures. Over the past two decades, hierarchical linear models (HLMs) have rapidly developed and been applied to model these structures. However, many datasets involve a cross-classified structure, instead of a purely hierarchical structure for which an HLM may not be appropriate. Given that cross-classified data structures are often ignored in many studies, this proposal intends to: (1) explore the consequences of ignoring a cross-classified structure in a cross-classified dataset with a dichotomous outcome; (2) examine differences due to estimation method; and (3) clarify the conditions under which there is a need to use the cross-classified random effects logistic model. Also, this proposal provides a demonstration of the cross-classified random effects logistic model and focuses on investigating possible recidivism predictors for incarcerated youth.
This methodological study will contribute to the literature on hierarchical generalized linear modeling and cross-classified random-effect logistic modeling. In addition, it will explore the importance of incarcerated youth’s reading ability and the effectiveness of an in-prison reading program on their recidivism. The effect of the county where the youth come from and the facility where the youth are retained on incarcerated youth’s recidivism will also be addressed.