Goldberg L R, Kilkowski J M
J Pers Soc Psychol. 1985 Jan;48(1):82-98. doi: 10.1037//0022-3514.48.1.82.
Subjects described themselves, using an alphabetically ordered list of 191 trait adjectives, which included sets of synonyms and antonyms, half of each type more difficult than the other half. Subjects were randomly assigned to one of two experimental conditions. In one condition, each adjective was listed with its dictionary definition; in the other condition, only the adjectives were listed. All subjects were administered a battery of demographic, cognitive, and personality measures. We analyzed both the relative consistency elicited by different pairs of terms and the individual differences in semantic consistency displayed by different sorts of subjects. Although the provision of definitions served to increase consistency (especially for the difficult antonyms), it did not decrease the range of consistency values across either synonym or antonym pairs. And, although interpair differences in semantic consistency were as difficult to predict in this study as in previous ones, individual differences were highly predictable. The implications of our many findings are discussed in the context of various hypotheses about semantic inconsistency in self-reports.
受试者使用一份按字母顺序排列的包含191个特质形容词的列表来描述自己，该列表包括同义词和反义词组，每种类型的一半比另一半更难。受试者被随机分配到两种实验条件之一。在一种条件下，每个形容词都列出了其词典定义；在另一种条件下，只列出形容词。所有受试者都接受了一系列人口统计学、认知和人格测量。我们分析了不同词对引发的相对一致性以及不同类型受试者在语义一致性方面的个体差异。尽管提供定义有助于提高一致性（尤其是对于难的反义词），但它并没有缩小同义词或反义词对之间的一致性值范围。而且，尽管在本研究中，语义一致性的词对间差异与之前的研究一样难以预测，但个体差异是高度可预测的。我们在关于自我报告中语义不一致的各种假设的背景下讨论了我们众多发现的含义。
