评价性条件反射效应:无条件刺激的呈现时长、效价强度与关联意识的作用

Evaluative Conditioning: the Role of Display Duration, Valence Intensity and Contingency Awareness

以真实商标图案为条件刺激,情绪图片为无条件刺激,探索无条件刺激呈现时长、效价强度与关联意识对评价性条件反射效应的影响。实验通过结合四图再认测验与基于项目分析,对关联意识的作用进行了详细探讨。结果表明,评价性条件反射效应只发生在无条件刺激长呈现水平与无条件刺激强效价水平;评价性条件反射效应的产生需要基于被试的关联意识。关联意识在呈现时长(效价强度)与评价性条件反射效应间的中介作用不显著。结果不支持评价性条件反射的内隐错误归因机制及联想-命题评价模型的相关论断,部分支持命题性解释模型。

Evaluative conditioning （EC） refers to the change of an attitude toward an affectively neutral object （conditioned stimulus, or CS）, following that the object＇s pairing with another positively or negatively valenced stimulus （unconditioned stimulus, or US）. EC is theoretically regarded as an associative learning process in associative-propositional evaluation （APE） model, but many controversies have arisen in empirical studies of EC. Some researchers found EC could not occurred without awareness of CS-US contingencies （supports propositional account）, but others didn＇t. It is also still unclear whether EC relies on much attention or not. Based on propositional account, the purpose of the present study was to investigate the effect of display duration of US, retest the effect of valence intensity, CS-US contingency awareness, and their integrated mechanism on EC, measured by explicit evaluative rating, combined With four-picture recognition test and item-based analyses. The hypotheses were tested in a sample of 122 college students （38 males）. In a 2（display duration） x 2 （valence intensity） x 2 （CS type） mixed design, with CS type as a within-group factor （CS＋ means that CS was paired with positive pictures and CS- means that CS was paired with negative pictures）, the participants were randomly assigned to four groups： US short display （120ms）-strong valence, US long display （1000ms）-strong valence, US short display-mild valence, and US long display-mild valence. The CSs were real pictorial trademarks but unfamiliar to the participants. USs were chosen partially from the International Affective Picture System and partially from the Internet. All pictures used in the experiment were selected based on pilot rating of valence. The experiment consisted of four sequential phases： the conditioning phase, evaluative rating phase, four-picture recognition phase and questionnaire phase. The results were analyzed by MANOVA on a global method of statistical analysis （participant-based）, and on the participants＇ responses to individual items （item-based） which was particularly for contingency awareness. Multiple regression analysis and chi-square test were also conducted. Results showed that evidence was obtained for EC effect on the global level, with more positive ratings of CS＋ than of the CS-. For display duration, EC emerged only on long display level, M （CS＋） = 4. 68, M （CS-） = 3. 86, F （1, 111） = 25. 19, p 〈 0. 001, η2 =0. 21. For valence intensity, EC was found only on strong valence 2 level, M （CS＋） = 4. 57, M （CS-） = 3.93, F （1, 111） = 14.57, p 〈 0. 001, ηp^2 =0. 16. Only the participants who were categorized as ＂contingency aware＂ showed significant EC effect. In addition, no effect of demand awareness and inference strategy was obtained. However, contingency awareness neither significantly mediated the relationship between display duration and EC effects, nor that between valence intensity and EC effects. This pattern of results partly supported the hypotheses, and emphasized the role of attention and awareness on EC effect. EC relies on US display duration, US valence intensity and CS-US contingency awareness, which in line with propositional account. Results also raise doubts about the views in implicit misattribution mechanism and APE model. EC is unlikely to be an associative evaluation process which requires little cognitive capacity and contingency awareness. Nonetheless, the insignificant mediation effect also indicates that propositional account needs to be further modified.