The Grade 1-9 curriculum policy has changed the traditional curriculum rationale and curriculum design. The large scale and profound influence of this change can be considered a revolutionary transformation during the educational process in Taiwan. The policy implementation has particularly focused on the elementary school and junior high school teachers, which involves not only the change in teachers¡¦ instructional patterns, but also teachers¡¦ professional roles as well as the transformation of their values. ¡@¡@Ideas of the Grade 1-9 curriculum policy have conveyed certain concepts. However, when the policy is implemented, a lot of details are involved, such as instructional plan, instructional method, action research, which require teachers¡¦ abilities for preparation, accompanying measures related to administration, or even the general public¡¦s and parent¡¦s questions about and support of the policy. All these have caused a lot of difficulties during the process of policy administration. Among them, the key point is whether the elementary school and junior high school teachers can identify themselves with the ideas of the curriculum reform, and dedicate themselves to implement this policy goal, which is closely tied to the substantial outcomes of this policy. Based on the context of this problem, it is worthy of a study on understanding this problem, and even resolving the problem. ¡@¡@An educational philosophy works as a basis to form a policy, and does not happen independently in the system of an educational structure. In turn, it is a response to the social context at that time, and is related to the social value system then. When the policies and ideas are implemented, a critical examination of social habits should not be ignored. In order to emphasize the policy analysis of curriculum reform implementation, if the society is considered the conceptual background of policy analysis, the policy formation can be more correctly and comprehensively understood, and can work as a basis to explain the problems due to the policy administration. ¡@¡@On these grounds, this study is based on the social concepts of the Grade 1-9 curriculum policy, and attempts to investigate the correlation between grassroots teachers¡¦ social values and the reform of the Grade 1-9 curriculum policy. The purpose is to explain the relationship between different teachers¡¦ social values and the policy of the curriculum reform, to understand cognitive differences as well as to test and verify the applicability of the methods to reduce the cognitive differences and resolve the cognitive conflicts. ¡@¡@In terms of the research method, the researcher first has referred to Inglehart¡¦s hierarchy of needs (1990:25) to measure the priority of the value index in the social values, and categories the teachers in the study into two different social value groups: materialism and post-materialism. The quasi-experimental method of the Social Judgment Theory (SJT) is applied to judge and analyze the study. The concept of ¡§V.C.S. Strategic Triangle, ¡¨ is applied to define the three indices, value-capacity-support (V.C.S.), in the Grade 1-9 curriculum policy implementation. The function and application of the lens model in the SJT, thus, are verified. It is hoped to test, verify and explain the different cognitive judgments on the beneficial results of the curriculum reform and V.C.S. policy implementation made by the teachers with different social values (criteria, cues, function form, weights, judgment policy and cognitive feedback). ¡@¡@Among the above-mentioned methods, the cognitive feedback model is especially applied to present the subjects¡¦ internal cognitive judgment on policy reference variables, for the purpose of helping the subjects to judge their level of cognitive change after they directly understand and compare their interpersonal learning with others. The findings are as follows: 1. Teachers with different social backgrounds and conditions have different value orientations. 2. Based on the SJT method, two groups with different social values have different cognitive levels of the curriculum policy reform¡¦s beneficial results. (1) The criteria of the curriculum policy reform¡¦s beneficial results among the subjects with different value orientations are very inconsistent, which indicates the cognitive differences between these two groups. (2) Groups with different social values cause different function form models and weights of the three policy reference variables: ¡§V¡¨ ¡§C¡¨ and ¡§S¡¨. Materialists tend to identify themselves with the variable, ¡§S¡¨ while the post-materialists with the other two variables, ¡§V¡¨ and ¡§C¡¨. (3) The three policy reference variables; namely, ¡§V¡¨ ¡§C¡¨ and ¡§S¡¨ among groups with different social values show lower similarity in the judgment policy. 3. After experiencing the interpersonal learning of cognitive feedback, groups with difference social values will make cognitive changes, which result in the narrowed cognitive differences between these two groups. (1) After the cognitive feedback, the cognitive consistency with the subjects will increase. (2) After the cognitive feedback, the actual value and estimated value of the test results among groups with different social values will have higher agreement. (3) After the cognitive feedback, the three function form, and weights, namely, ¡§V¡¨ ¡§C¡¨ and ¡§S¡¨ among groups with different social values have been adjusted toward the cognitive orientation different from theirs (4) After the cognitive feedback, the three cues, namely, ¡§V¡¨ ¡§C¡¨ and ¡§S¡¨ among groups with different social values show higher similarity in the judgment policy. 4. The results of the questionnaire found that groups with different social values and social attributes (such as social conditions during the growing process, family¡¦s social-economic status, age and educational background) have different levels of identification with the three indices, namely, ¡§V¡¨ ¡§C¡¨ and ¡§S¡¨ Materialists tend to identify themselves with the variable, ¡§support,¡¨ while post-materialists with the other two variables, ¡§V¡¨ and ¡§C¡¨. The research results are identical with the research results of the SJT.