A longitudinal study of political attitudes and behavior of high school students reveals varying results when two different methods of data analysis are applied. A sample of 293 students was questioned (during their sophomore year in 1974 and again in 1976 when they were seniors) about political orientation, number of social studies classes taken, social studies classroom climate, school activities, and gender. The purpose of the study was to expand upon earlier research (Beck and Jennings, 1982) linking high school activity with youth civic orientation. When path analysis was applied, the study replicated Beck and Jennings' findings, which also resulted from path analysis. The contribution of social studies classes to political attitudes was found not to be significant while an open social studies classroom environment led to positive political attitudes. However, when two-stage least squares analysis was applied, classroom climate was found not to be a significant predictor. Also, while prior political attitudes led to political participation, increased participation resulted in more negative attitudes. For various reasons, the two-stage least squares method appears to be the most reliable, least restrictive technique and thus more appropriate for analyzing complex relationships. (KC)