|
|||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
|
between themselves, indicating a high degree of reliability between these expert judgments from opposing blocs. If our scales are valid in the predictive or criterion validity sense, they should correlate in certain ways with the expert ratings. Specifically, we expect very high correlations with our Marxist scale, for the U.S. and Soviet ratings on their face are Marxist in character. Indeed, the experts ratings ought to correlate higher with our Marxist scale than the equally reliable economic leftism sub-scale. Because our liberalism scale measures a dimension of party ideology that is entirely unrelated to our Marxism factor, there should be little or no relationship between the experts ratings and liberalism. Results of the comparisons are given in Table 14.5.
The correlations in Table 14.5 unequivocally support the expectations. For both sets of expert ratings, the correlations are highest with the Marxism scale, slightly less for the economic leftism scale, and virtually nonexistent for the liberalism scale. The fact that the relevant correlations are higher with U.S. than Soviet experts is consistent with the greater variance for the U.S. expert ratings (a four-point scale) than the Soviet ratings (a three-point scale). The results are completely as expected. The goal orientation cluster involved 33 variables, far more than any other section of the conceptual framework, organized to support a distinction between party strategy--the basic plan for winning government office--and party tactics--the specific activities of the party in light of its strategy. Three basic party strategies were identified as (1) open competition, (2) restricting competition from other parties, and (3) subverting the political system. Each party was scored separately for its strategy according to its overall role in the political system. In addition, a series of tactics was posited as directly related to each of these strategies. Moreover, other sets of tactics were identified as indirectly related to any strategy. The direct and indirect tactics together accounted for 30 of the 33 variables, the other 3 being the key strategy variables. As research progressed on the project, it became clear that the available funds would not be adequate to collect all the data for all the parties in the study. The goal orientation cluster, with its numerous variables, was a place where major economies could be realized while isolating the information loss to only one conceptual domain. We decided to continue coding all pasties for the three strategy variables but to cease coding parties for the direct tactics and certain of the indirect tactics, stopping with less than 30 percent of the parties coded on 21 of the 33 variables in the cluster. These data losses were regrettable but necessary to move the project toward completion with dwindling resources. Moreover, the direct tactics were keyed to the strategy variables, which were retained, and the indirect tactics that were eliminated were reflected in another variable in the "autonomy" cluster (described in the next section). All parties in the study were scored for these strategy variables: 6.00 Open Competition. The higher the score, the greater the reliance on electoral competition for winning government office. Approximately 45 percent of the parties followed a "pure" strategy of open competition, 10 percent relied exclusively on restricting competition, and only about 2 percent were dedicated to subversion. The remaining parties employed some mix of strategies, with the most common combination being open competition with some degree of restriction. Validation of our coding of party strategies can be approached by studying the posited relationships between strategies and tactics. Unfortunately, this approach is impaired by the large proportion of missing data for the direct tactics due to the curtailment of our research for these variables. Nevertheless, the limited evidence available offers some grounds for validation, although the matter cannot be pushed very hard. The tactics thought to be directly associated with certain strategies are listed in Table 14.6, along with the correlations between each tactic and its associated strategy as calculated for approximately 25 percent of the parties in the study. As shown in Table 14.6, only three of the sixteen posited relationships fall short of significance at the .01 level. All three of these are tactics thought to be directly related to open competition. According to our limited data base, advertising candidates by signs and posters, promoting candidates through direct contact, and holding public rallies are not particularly associated with a strategy of open competition. While these findings might question the validity of our scorings of party strategy, they also question the initial identification of |
|||||||||||||||||||||||||||