Novy D M
Department of Anesthesiology, University of Texas Medical School, Houston 77030.
J Clin Psychol. 1992 May;48(3):308-14.
This study investigated the reliabilities and conjectured factor structure of the subscales comprised by Lorr and DeJong's (1986) abbreviated form of the Interpersonal Style Inventory (ISI). The results suggest that internal consistency reliabilities of the subscales are presently inadequate for clinical or research purposes. The five-factor structure described by Lorr (1986) appeared to conform well to the empirical grouping in the data set. Only one scale, Approval Seeking, appeared incorrectly placed according to the a priori description. Another weakness in the a priori structure was that the Rule-Free subscale had only a minor role in identifying its conjectured factor. These results were based on ISI protocols from 268 adults. Overall's (1974) marker variable factor analysis was implemented to assess the adequacy of the a priori factor structure.
本研究调查了洛尔和德容（1986年）编制的人际风格量表（ISI）简版所包含的分量表的信度及推测的因子结构。结果表明，目前这些分量表的内部一致性信度对于临床或研究目的而言是不足的。洛尔（1986年）描述的五因素结构似乎与数据集中的实证分组非常吻合。根据先验描述，只有一个量表，即寻求认可量表，似乎放置不当。先验结构的另一个弱点是，无规则分量表在识别其推测因子方面只起了次要作用。这些结果基于268名成年人的ISI测试记录。采用奥弗尔（1974年）的标记变量因子分析来评估先验因子结构的充分性。
