Psychometric assessment of social cognitive tasks
Psychometric assessment of social cognitive tasks
复制标题
DOI：
10.1080/23279095.2020.1807348
复制
发表时间：
2020-08-22
影响因子：
1.7
通讯作者：
Scherzer, Peter B.
中科院分区：
文献类型：
--
作者：
Gourlay, Catherine;Collin, Pascal;Scherzer, Peter B.
虽然近年来对社会认知（SC）的兴趣显著增加，但与用于测量其组成部分的许多任务相关的心理测量数据在健康人群中仍然有限，迄今为止只有5篇文章发表。因此，现在谈论关于具体组成部分或组成部分的最佳测试以及可能的文化差异的共识还为时过早。本研究试图部分填补这一空白，检查电池的SC任务的心理测量学特性在100个健康的成年人，年龄在18-85岁的样本。最初，九个任务评估四个SC组件被选中：情绪识别，心理理论，归因偏差和社会判断。结构效度和效标关联效度采用因素分析和相关分析进行评估。还调查了不同年龄和性别组的表现。通过内部一致性、评分者间和编码者间一致性评估可靠性。结果表明，令人满意的性能的Ambibliosis意图敌对情绪责备分数，社会判断任务，面部情绪识别测试，和一个修改版本的奇怪的故事任务。在考虑人口统计学和认知因素后，发现各组之间在年龄和性别方面存在统计学显著差异。然而，这些措施与关系质量的相关性大多非常低，提出了他们的伴随有效性的问题。其他任务显示出次优特性，这表明一些常用的测试需要进一步验证或修改，以确保研究结果的质量。根据研究结果，建议的措施，为今后的研究和局限性进行了讨论。
Although there has been a marked increase in interest in social cognition (SC) in recent years, psychometric data relating to many tasks used to measure its components remain limited in healthy populations with only five articles published to date. It is accordingly premature to speak of a consensus concerning the specific components, or best tests of the components, and possible cultural differences. The present study sought to partially fill that gap, examining the psychometric properties of a battery of SC tasks in a sample of 100 healthy adults aged 18-85 years old. Initially, nine tasks assessing four SC components were selected: emotion recognition, theory of mind, attributional bias, and social judgment. Construct validity and criterion-related validity were assessed using factor and correlational analyses. Performance across age and sex groups was also investigated. Reliability was assessed through internal consistency, interrater and intercoder agreement. Results indicated satisfactory properties for the Ambiguous Intentions Hostility Questionnaire-blame score, the Social Judgment Task, the Facial Emotions Recognition Test, and a modified version of the Strange Stories Task. Statistically significant differences were found between the groups with regard to age and sex after accounting for demographic and cognitive factors. However, the correlations of these measures with relationship quality were mostly very low, raising questions about their concomitant validity. Other tasks showed sub-optimal properties, suggesting that some frequently used tests require further validation or modifications to ensure the quality of research findings. Based on the results, recommended measures for future studies and limitations are discussed.