矛盾态度心理机制的理论模型述评

A Review on Theoretical Models About Psychological Mechanisms of Ambivalent Attitude

矛盾态度已发展为态度研究领域一个重要方面,在理论和实证研究都取得重要进展,但是对矛盾态度心理机制的探究还处于滞后的态势。本研究试图对矛盾态度心理机制研究做一个较系统的梳理,引入对象与评价的联结模型、态度元认知模型、重复加工模型和表征分布式联结主义模型等四种态度理论模型有关解析,并结合相关实证研究予以阐述和评价。未来该领域的研究应加强矛盾态度研究的元分析、各研究之间的沟通和整合、实证研究范畴的拓展、新技术的引入等问题的研究。

On the psychological mechanisms of ambivalent attitude （AA）, Cacioppo and Berntson advanced the model of evaluation spacein 1994, and Priester and Petty proposed the GTM （The gradual threshold model） in 1996, both of which focus on the formation of AA and form the theoretical basis of the internal mechanism of AA, but there is a lack of studies in such respects as the representation and resolution of AA, and its impact on information processing. With more thorough research on AA, there has been a growing need for the guidance of theories on psychological mechanism. This article attempts to do some generalization and appraisal of the theoretical models of the psychological mechanism of AA. Moreover, the current progress in attitude theories and empirical studies provides a possible method for in-depth explanation of the internal mechanism. By virtue of literature research, the article introduces four new theoretical models associated with the analysis of psychological mechanisms underlying AA, i.e. the model of attitude as object-evaluation association, the meta-coguitive model of attitude, the iterative reprocessing model, the model of distributed connectionist representation and so on, and conbined recent empirical findings with our analysis and evaluation. For different theoretical hypothesis and construction, the four models on attitude focus differently on the representation, formation and resolution of AN. The object-evaluation association model stresses the attitude formation process to analyze how ambivalence generates, provides a novel explanation of AN from the scenarios interaction and the diversity of classification, and holds that ambivalence is not caused by two kinds of evaluation activated spontaneously, but by the rapid conversion among different assessment classifications. The meta-cognitive model of attitude pays more attention to the analysis of how ambivalence is represented in the memory, and classifies AN into two forms from the angle of meta-coguition of AN： explicit and implicit, and also makes some specifications of explicit and implicit conversions concerning attitude change. The iterative reproeessing model emphasizes the role of the analysis of higher cognition processing in solving ambivalence, and furthers the studies of the corresponding neuron-physiological mechanism. The model of distributed connectionist representation lays its emphasis on building from the momentary scene in analyzing the activation and representation of AA, and weakens the difference of characterization between the ambivalent and non-ambivalent attitudes. On balance, the paper prol^ses an integrated view that the theories of the psychological mechanism of AA should be integrated, through which we could understand the psychological mechanism of AA more thoroughly. In short, although the above studies on AA can help us understand from a better angle the structure of attitudes, attitudes impact on information processing, and attitude stability versus attitudes malleability, there exist many deficiencies in the current research in theoretical and empirical aspects; future research should pay attention to the meta-analysis of AA, the exchange and integration of researches, expanding of empirical research category and introduction of new technologies.