Wiggins J S, Trapnell P, Phillips N
Multivariate Behav Res. 1988 Oct 1;23(4):517-30. doi: 10.1207/s15327906mbr2304_8.
Previous research has suggested that the Interpersonal Adjective Scales (IAS) have properties that are compatible with the substantive, structural, and external characteristics required by the circumplex-based interpersonal theory which guided scale construction procedures. In the present study, we describe an item-analytic procedure that identifies and selects items in terms of their estimated geometric location within a circumplex model and we apply the procedure to the task of reducing the 128-item IAS to a 64-item short form version (IAS-R). Scales constructed to conform to a circumplex model are evaluated by somewhat different structural criteria than are those developed to conform to other multivariate models and these differences are illustrated. The resultant IAS-R was found to have improved substantive and structural characteristics and acceptable reliability. We discuss possible applications of this highly efficient measure of interpersonal behavior.
先前的研究表明,人际形容词量表(IAS)具有一些特性,这些特性与基于环形模型的人际理论所要求的实质、结构和外部特征相兼容,该理论指导了量表的构建过程。在本研究中,我们描述了一种项目分析程序,该程序根据项目在环形模型中的估计几何位置来识别和选择项目,并将该程序应用于将128个项目的IAS缩减为64个项目的简短形式版本(IAS-R)的任务。为符合环形模型而构建的量表,其评估的结构标准与为符合其他多变量模型而开发的量表有所不同,本文对此差异进行了说明。结果发现,由此产生的IAS-R具有更好的实质和结构特征以及可接受的信度。我们讨论了这种高效的人际行为测量方法的可能应用。