Probhems and aims.: Generally speaking, there are two aspects in the jud gement of items of a social attitude scale.The one (S) is that in which the items are judged in terms of agreements with the statements of the attitude scale, and the other (S´) is that in which they are judged in terms of oppositions to the statements.When examined statistically, S-S'has a significant difference which was d emonstrated byour previous researches.And then, why is the difference significant will be able to point out, as one of the chief reasons, that the mechanism in the judgement of S is different from that of S.Precisely, attitudes are always related to definite stimuli or stimulus situations.(subject-object relationship).The concept of attitudes are often de noted in termsof its objects, but the content of an attitude is decided by the property of the subjecto bject relationship which has established at that time, and which has fulfilled the important role as a parameter. Therefore, in order to explain the bi-polarity of social attitude, we will adopt the mechanism of value and the mechanism of radicalism-conservatism, as examples of subject-object relationship, and examine them experimentally. Procedure of experiment.: We must, for the first time, construct attitude scales, each statement of which is emphasizing only one of the values, or raidicalism, or conservatism. (1) We have established, for the first time, attitude scale toward “Labor” by the method of equal- appearing intervals, which consisted of80statements, and the value type of each statement was decided and classified by students'objective ratings into six types: theoretical, social, aesthetic, religious, economical, and political.If we research statistically the frequencies of the value type of each statement toward which the responses of “pro” or “anti” were made, we will be able to decide whetherthe responses are made through homogeneous value-mechanism or through heterogeneous valuemechanism. When the value-type of each statement toward which the bi-polaric responses are made, are not the same but different, we will be able to consider that an individual made responses through heterogeneous value-mechanism.From such a viewpoint as this, we have asked285students to make such bi-polaric responses only one time for one scale. (2) We have made a social attitude scale which consisted of14radical and8conservative statements out of Eysenck's inventory.We presented it to the same students, asking them to make bi-polaric responses only one time for one scale.The procedure of this experiment is the same as that in the case of value-mechanism described above. Results.: (1) When one make his bi-polaric responsestoward the items of an attitude scale, which were certified for emphasizing one of six types of value, or radicalism or conservatism, he is used to make his responses, very often, through heterogeneous mechanism, and very few, through homogeneous mechanism. (2) When one makes his bi-polaric responses toward each item of an attitude scale, there is a significant difference between one pole (S) and the other (S´).Why is the difference significant.We will be able to point out the above conclusion (1) as one of its chief reasons.That is to say, it is because the mechanism in the judgement of S is differ statistically from that of S´.