A Causal Model of Positive Health Practices

This study supports the position that causal models developed a priori preclude replication with varied samples. Based on a critique of a study of positive health practices among adults (Muhlenkamp & Sayles, 1986), a causal model of positive health practices for adolescents was developed a priori from a theoretical formulation. Using data from a sample of 165 adolescents who responded to the Personal Lifestyle Questionnaire, the Rosenberg Self-Esteem Scale, Part 2 of the Personal Resource Questionnaire which measures a social support system, and a demographic data sheet, the intercorrelations among the study variables were analyzed using correlation coefficients. The causal model was then tested with the adolescent data using the LISREL VI program. The results showed a relatively good fit of the model to the data via a number of indicators. The model was then applied to data published from adults (Muhlenkamp & Sayles) using the LISREL VI program. The results indicated that there was a relatively poor fit of the model to the adult data, thus demonstrating the problem of replicating causal models with varied samples when the correct approach to causal modeling is used. The discussion focuses on theoretical and methodological reasons for the findings.