58 absent from this more elaborated breakdown. This means that the weakest learning occurred among the middle-class blacks excluded from this particuThis particular finding will remain undiscussed without fur lar analysis. ther supportive research. In addition to these interactions on general subject characteristics, conditional associations were also examined at high and low levels of ten politically relevant antecedent variables, all measured at the first test session, before any shows were seen. For each of these continuous variables, the overall sample was split near the median into two groups. This allows analysis of the relative impact of the legislature programs on subjects who differed in original knowledge, interest, efficacy, attitude, and communication behavior. The numerical values of the high vs. low classifications on the six most important variables are indicated in Table 6. The following paragraphs discuss the findings for each antecedent test variable. Initial Knowledge. There were surprisingly small differences between the knowledge acquisition of the well-informed and poorly-informed subjects (Table 6a). Those who originally scored low on the general knowledge index learned slightly more than the more knowledgeable subjects. Since this occurred on the structural knowledge index as well as the general knowledge index, a ceiling effect explanation seems unlikely. The well-informed group showed the only increase in political efficacy, while experimental exposure was not related to efficacy among the poorly-informed subjects. On the other hand, the less knowledgeable group became more favorable toward the legislators, while the evaluations of the more knowledgeable subjects were not affected by the experiment. Table 6a Conditional Associations between Experimental Treatment and Dependent Variables, at High and Low Initial Levels of Political Knowledge and Interest Low Knowledge = Score of 0-2 on Basic Political Knowledge Index at Tl. Low Interest = Response of "Not" or "Little" on Interest in Legislature item at T1. High Interest = Response of "Somewhat" or Very" on Interest in Legislature item at T1. 60 Initial Interest. Those who were previously interested in the legislature further increased their interest due to the treatment, while the low interest group was little stimulated (Table 6a). Otherwise, the low inter est and high interest students did not differ in their responses to the program series. Initial Evaluation. Those who were originally least favorable in their evaluation of the legislators were the only ones who were positively affected by the programs, possibly because they had the most room for favorable change (Table 6b). They became more favorable toward the legislators, and increased in perceived political efficacy. However, the initially favorable subjects tended to acquire more knowledge. Initial Efficacy. Subjects scoring high on the first efficacy index were the only ones stimulated to become more efficacious; the low efficacy subjects showed no effect on this variable (Table 6b). The former also became more interested in the legislature. Initial Discussion. In general, those who previously discussed politics with parents and peers were affected more than non-talkers on the political interest variable (Table 6c). Other differences were very inconsistent, and interactions on untabled variables were not hypothesized. On the last four antecedent factors, only a few interaction analyses were expected. Thus, the conditional associations between experimental treatment and dependent variables are described in the text rather than presented in tabular form. Initial Viewing Plans. One-fourth of the students said at Tl that they expected to watch legislature programs at home. This group was only slightly influenced by the experimental treatment to watch at home after Table 6b Conditional Associations between Experimental Treatment and Dependent Variables, at High and Low Initial Levels of Efficacy and Attitude toward Legislators Low Efficacy = Score of 0-3 on Political Efficacy Index at Tl. Low Attitude = Score of 0-4 on Attitude toward Legislators Index at Tl |