LISREL Analysis is suggested as a method of modeling relationships among variables to determine causality. LISREL is a computer program that uses a maximum-likelihood procedure to analyze structural equations. The variables in the tested hypotheses can be both directly observed variables and latent variables. LISREL was applied to the findings of a study in drug education research in order to provide a model. The constructed LISREL model was contrasted with the ANOVA-based model. The LISREL had greater explanatory value for dealing with antecedent variables present in complex human health behaviors.