ITEM CONTEXT EFFECTS ON PERSONALITY-SCALES - MEASURING CHANGES THE MEASURE
ITEM CONTEXT EFFECTS ON PERSONALITY-SCALES - MEASURING CHANGES THE MEASURE
复制标题
DOI：
10.1037/0022-3514.55.2.312
复制
发表时间：
1988-08-01
影响因子：
7.6
通讯作者：
KNOWLES, ES
中科院分区：
文献类型：
--
作者：
KNOWLES, ES
重复考虑单一特质人格测验的项目应该增加测试相关图式的可及性、清晰度和内容。因此，项目答案应该变得更加两极化，一致性和可靠性作为以前回答的项目数量的函数。两项研究未混淆项目内容从项目序列的位置，使变化，由于以前的项目的数量可以进行评估。在研究1中，120名大学生填写了James（1957）的内部-外部控制点（I-E）量表的平衡表格。在研究2中，600名本科生填写了I-E量表、教条主义量表、泰勒明显焦虑量表或马洛-克劳社会期望量表。在这两项研究中，一个项目与测试其余部分的相关性随着其在测试中的序列位置线性增加。在对I-E数据的重新分析中，研究3表明，答案变得更加两极化，特别是对于内部，并且在测试后期变化较小。测试和信息处理的影响进行了讨论。
Repeated consideration of items on single-trait personality tests should increase the accessibility, clarity, and content of test-related schema. Consequently, item answers should become more polarized, consistent, and reliable as a function of the number of previous items answered. Two studies unconfounded item content from item serial position so that changes due to the number of previous items could be assessed. In Study 1, 120 undergraduates filled out counterbalanced forms of James's (1957) internal—external locus of control (I—E) scale. In Study 2, 600 undergraduates filled out the I—E scale, Dogmatism Scale, Taylor Manifest Anxiety Scale, or Marlowe-Crowne Social Desirability Scale. In both studies, an item's correlation with the rest of the test increased linearly with its serial position in a test. In a reanalysis of the I—E data, Study 3 showed that answers became more polarized, especially for internals, and less variable later in the test. Implications for testing and information processing are discussed.