【作者】 田录梅；

【导师】 张向葵；

【作者基本信息】 东北师范大学 ， 发展与教育心理学， 2007， 博士

【摘要】 自尊水平与心理健康尤其是情感健康密切相关,高水平自尊还具有很强的自我服务效应。但自尊不仅有水平高低之分,还有不同类型之别,而且自尊的认知加工偏好不仅有认知加工方式上的自我服务偏好,也有认知加工内容方面的加工偏好,这种加工偏好与情感健康之间的关系目前尚存在较大争议。那么,不同自尊者对自我相关信息的认知加工偏好尤其是认知加工内容方面的偏好是否有明显差异?这种可能不同的认知加工偏好对个体的情感健康是否会产生不同影响?基于上述思考,本研究主要采用4(自尊类型)×2(自我信息效价:积极/消极)和3(认知加工偏好类型)×2(情感测量时间:加工偏好前/加工偏好后)混合设计,通过问卷调查和实验室研究相结合探讨了不同自尊类型的被试对自我相关信息的认知加工偏好差异,以及这种可能不同的认知加工偏好是否对其情感反应(情感健康的状态性指标,便于操作)亦有不同的影响。这也是研究1和研究2要解决的两个主要问题。在正式研究之前先通过预备研究选取记忆测验的材料以用来测量被试的记忆偏好(认知加工偏好的一个指标)。然后以大学生为被试根据其外显自尊水平(由Rosenberg(1965)自尊量表测得)和内隐自尊水平(由IAT测得)一致性的四种关系划分出四种自尊类型:安全高自尊、脆弱高自尊、一致低自尊和不一致低自尊,考察这四组被试对自我相关信息的回忆偏好(研究1),并将其加工偏好分为积极加工偏好、消极加工偏好和无偏好三类,接着考察不同加工偏好对被试此后的情感反应(由PANAS-R测得)是否有不同影响(研究2)。鉴于有研究者指出,自尊发挥作用最大之处是在消极事件之后,自我威胁会扩大自我增强偏好,因此,研究3和研究4继续探讨了不同自尊者在诱导的失败后的认知加工偏好及其对情感反应的影响以及这些效应分别与研究1和研究2相比是否有增强的趋势。研究5和6则通过控制被试加工每条自我相关信息的时间(500ms)进一步探讨了不同自尊者的认知加工偏好是否是一种策略以及控制这种可能的策略使用后认知加工偏好对情感反应的影响是否消失。最后一个研究(研究7)通过对研究1的被试进行为期一学期的追踪,考察不同自尊者的认知加工偏好对其情感反应的影响能否持久从而对其稳定的情感健康产生贡献。最后,基于这7个基本呈递进关系的实证研究之结果,本研究可得出如下几点结论:(1)由于有相当一部分被试外显自尊和内隐自尊水平高低不一致,且心理后果(情感反应)也有异于其他被试,因此,高自尊异质性问题的确存在。(2)无论自尊水平还是自尊类型,不同自尊者在对自我相关信息进行加工时有不同的认知加工偏好,而且这种不同加工偏好对情感反应有不同影响。具体地说,就组间差异而言,脆弱高自尊者比其他类型自尊的被试有更积极的认知加工偏好,而积极的加工偏好显然引发了更少的消极情感反应;就组内差异而言,任何类型自尊的被试都有相对于消极信息的积极信息加工偏好,而这种偏好能缓冲或减少被试的消极情感反应。(3)不同自尊者在受到失败的自我威胁后,其积极加工偏好及其对情感反应的影响均有增强的趋势。具体地说,就组间差异而言,脆弱高自尊者比两种低自尊者都有更强的积极加工偏好,安全高自尊者也出现了相对于一致低自尊者的积极加工偏好,而积极加工偏好不仅引发了更少的消极情感反应也引发了更多的积极情感反应;就组内差异而言,除一致低自尊者外,其余被试都有相对于消极信息的积极信息加工偏好,而积极信息加工偏好不仅减少了被试的消极情感反应,也增加了被试的积极情感反应。(4)不同自尊者在加工信息的时间受限(500ms)时,其加工偏好及其对情感反应的影响组间差异消失,但组内加工偏好及其对情感反应的影响依然存在,说明作为组间差异存在的认知加工偏好是一种策略,而作为组内差异存在的加工偏好可能是自动化的结果,人们也很可能因此而获得相对的心理健康。(5)不同自尊者在加工自我相关信息上的不同认知加工偏好对情感反应的影响是暂时的,并不能持久。更多还原

【Abstract】 Levels of self-esteem are closely correlated with mental health, esp. with emotional health. High self-esteem has strong self-serving effect too. Self-esteem not only has division of high and low level, however, it also has many different types, and the cognitive processing biases of self-esteem include not only biases in cognitive processing ways and also biases in cognitive processing content, whose relationship with emotional health has been being disputed. Then, are there obvious differences in cognitive processing bias of different people with different self-esteem, esp. in bias of processing content? Can these possible different processing biases affect differently people’s emotional health?Based on these consideration, mixed designs of 4(types of self-esteem)×2(valence of self-relevant information: positive/negative) and 3(types of cognitive processing bias)×2(time of affect measuring: before processing bias/after processing bias) were adopted in this dissertation study. They explored differences of biases in processing self-relevant information of different people with different self-esteem types and whether these possible different processing biases could also affect differently people’s affective responses (an index of state emotional health) , which were the main questions solved in Study 1 and Study2, by way of combining questionnaires and experiments. A pilot study to choose material of memory test was done before formal studies in order to measure participants’memory bias (an index of cognitive processing bias). Then four types of self-esteem of undergraduate-participants were differentiated according to the four combinations of levels of explicit self-esteem (measured by Rosenberg (1965)’s Self-Esteem Scale) and implicit self-esteem (measured by IAT): secure high self-esteem, fragile high self-esteem, congruent low self-esteem and discrepant low self-esteem, in order to investigate the memory bias in self-relevant information among these four types (Study 1). And three types of cognitive processing bias were differentiated: positive processing bias, negative processing bias and without bias, in order to further explore whether they affected the participants’affective responses (measured by PANAS-R) after using them (Study 2). Because some researchers had pointed out that the most effect of self-esteem happened after negative events and self-threats could magnify self-enhanced biases, Study 3 and Study 4 continued to explore cognitive processing biases of the participants with different self-esteem after induced failure and their effects on affective responses and to explore whether there were enhanced trends compared to Study 1 and Study 2 respectively. Study 5 and Study 6 further explored whether the cognitive processing biases of the participants with different self-esteem were only a kind of strategy by controlling the time (500ms) of processing each piece of self-relevant information and whether the effects on affective responses of cognitive processing biases disappeared by controlling the possible strategy. The last study (Study 7) investigated whether this kind of influences on affective responses could last longer to contribute to stable emotional health by tracing the participants in Study 1 for a semester.Finally, based on the results of these 7 empirical researches step by step, the following conclusions can be made in this dissertation study:(1) Because many participants’explicit and implicit self-esteem were discrepant and their mental consequences (affective responses) were also different from other participants, the heterogeneity of high self-esteem does exist.(2) Participants with different self-esteem, whether different levels or different types, had different biases in processing self-relevant information, which also had different effects on their affective responses. In detail, concerning the differences among the four self-esteem groups, participants with fragile high self-esteem had more positive processing bias than other participants, and positive processing bias produced less negative affect; concerning differences in each group, anyone had positive-information processing bias relatively to negative information, and this bias buffered or reduced the participants’negative affect.(3) There were enhanced trends of positive processing bias and its effects on affective responses of people with different self-esteem after the participants had experienced self-threat of failure. In detail, concerning the differences among the four self-esteem groups, participants with fragile high self-esteem had more positive processing bias than those with low self-esteem, and those with secure high self-esteem had also positive processing bias relatively to those with congruent low self-esteem. Positive processing bias produced less negative affect and more positive affect. Concerning the differences in each group, the participants had positive-information processing bias relatively to negative information exclude those with congruent low self-esteem, and this positive processing bias reduced negative affect and increased positive affect.(4) The differences of processing bias and its effects on affective responses among the four groups disappeared when the participants processed the self-relevant information in limit time of 500ms, but the differences in each group still existed, which suggested that the cognitive processing bias as a kind of difference among different groups is a kind of strategy and the cognitive processing bias as a kind of difference in each group is possibly an automatic result and people get relative mental health consequently.(5) The effects on affective responses of different cognitive processing biases of the participants with different self-esteem were temporary and couldn’t last long.更多还原