a Department of Psychology , University of Tennessee at Chattanooga.
b Department of Psychology , Illinois Institute of Technology.
J Pers Assess. 2018 Jul-Aug;100(4):375-388. doi: 10.1080/00223891.2017.1406362. Epub 2018 Jan 10.
Exploratory bifactor models with keying factors were applied to item response data for the NEO-FFI-3 and HEXACO-PI-R questionnaires. Loadings on a general factor and positive and negative keying factors correlated with independent estimates of item valence, suggesting that item valence influences responses to these questionnaires. Correlations between personality domain scores and measures of self-esteem, depression, and positive and negative affect were all reduced significantly when the influence of evaluative content represented by the general and keying factors was removed. Findings support the need to model personality inventories in ways that capture reactions to evaluative item content.
探索性双因素模型与关键因素被应用于 NEO-FFI-3 和 HEXACO-PI-R 问卷的项目反应数据。一般因素和积极与消极关键因素的负荷与项目效价的独立估计相关，这表明项目效价会影响对这些问卷的反应。当去除一般因素和关键因素所代表的评价内容的影响时，人格领域得分与自尊、抑郁、积极和消极情绪的测量之间的相关性显著降低。研究结果支持需要以捕捉对评价项目内容的反应的方式来对人格量表进行建模。
