CAUSAL MODELS OF VIOLENCE BEHAVIORS OF VOCATIONAL STUDENTS: VALIDATION OF EFFECTS FROM LONGITUDINAL AND CROSS-SECTIONAL STUDIES

Abstract

This research aims 1) to study the causal model of the violence behavior of vocational students from longitudinal research plans and cross-sectional research plans and 2) to compare the accuracy of a causal model of violence behavior of vocational students from a longitudinal research plans and cross-sectional research plans. The sample in this study was vocational students who were studying for vocational certificate and high vocational certificate. The samples came from multi-stage random sample of 1,049 people and the collection of data was repeated 3 times. The instrument in this study was a questionnaire of the violence behavior of vocational students. The researcher developed the instrument based on the theory of planned behavior and the accuracy analysis of the causal model of the violence behavior of vocational students. The theory of planned behavior that the researcher used was from longitudinal research plans and cross-sectional research plans with a structural equation model (Structural Equation Modeling: SEM). The result showed that     1) The causal model of violence behavior of vocational students from longitudinal research plans had the proportion of the variance in behavior latent of using violence was explained by the independent variables of approximately 4.60 percent , 19.40 and 33.00 respectively. The cross-sectional research plans had the proportion of the variance in behavior latent of using violence was explained by the independent variable percentage of 17.20, 25.20 and 36.70 respectively.

            2)         The structural equation modeling from longitudinal research plans could explain the violent behavior of students in vocational education more consistent with the empirical data than the structural equation modeling from the cross-sectional research plans.