政治意識構造論の試み(その1)

抄録

Various attitudinal factors are considered as media connecting demographic factors and socio-economic conditions to voting behavior. However, the structure or interrelatedness among these attitudinal factors themselves have not been sufficiently clarified.
A cluster of attitudes may at times form an hierarchical structure by a frame of reference, or, while in conflict to each other, they somehow preserve a balance and form so diffuse an aggregate that it can not be called an attitude structure in the narrow sense of the term. However some pattern of interrelatedness should be found among these attitudes. Here the term “attitude structure” is used in such a wider sense, and the object of this paper is to suggest an empirical model of such an attitude structure.
The starting point of this study is to establish the kind of attitudinal factors which constitute an attitude structure, and it is best to establish them based on a theoretical hypothesis concerning attitude structure. It is usually said that a certain attitude is composed of an affective component. When a cluster of attitudes constitutes a structure, the affective components and the cognitive components of the various attitudes are mutually correlated, respectively constituting an affective structure and a cognitive structure. An affective structure is further divided into a structure of the direction of affection and a structure of the intensity of affection. These three structures can be classified into some levels, for instance by a criterion whether they are nearer political behavior or nearer social conditions.
Selecting variables which represent each structure and each level in the above mentioned theoretical chart, and after a series of correlation analyses among 2 or 3 such variables, an empirical model of attitude structure has been formed, revising a part of the theoretical chart mentioned above. Further, in order to support this model, multi-variate analysis, in this case, factor analysis was performed. As its consequence, 5 factors were found, indicating upper and lower levels of the affective structure, upper and lower levels of the cognitive structure and a structure of the intensity of affection.
Lastly, sub-models were examined. As the political culture of Japan is pluralistic, this model may be considered as an aggregate of several sub-models different in character. Therefore, by dividing the samples into two groups according to principal attitudinal factors, we examined how the types of the factors which were just discovered would change, and no great difference has been found among the sub-groups.