Bagby R M, Parker J D, Taylor G J
Department of Psychology, Clarke Institute of Psychiatry, Toronto, Ontario.
J Pers Assess. 1991 Apr;56(2):238-53. doi: 10.1207/s15327752jpa5602_5.
In the past decade, alexithymia has emerged as a heuristically useful personality construct used to explain the pathogenesis of a variety of physical illnesses, including classical psychosomatic diseases, somatization disorders, hypochondriasis, and somatoform pain disorders. Unfortunately, research evaluating the alexithymia construct has been conducted with little attention to assessing the psychometric properties of various scales used to measure it. In two separate studies, we examined various scale and item properties as well as the factor structure and validity of the Minnesota Multiphasic Personality Inventory Alexithymia Scale (MMPI-A), one of the most commonly used scales to assess alexithymia. In Study 1, the 22 items that comprise the MMPI-A were extracted from a computerized MMPI data bank which included separate samples of psychiatric inpatients and outpatients. Poor item-to-scale characteristics and only moderate levels of internal reliability were found for both samples. Factor analysis produced factors that were poorly related to the theoretical domains of the alexithymia construct. In Study 2, we found little support for validity of the scale as those patients identified as alexithymic and nonalexithymic by the MMPI-A did not differ on several theoretically relevant scales. These results question seriously the value of the MMPI-A in investigating the alexithymia construct.
在过去十年中,述情障碍已成为一种在启发式研究中有用的人格结构,用于解释多种身体疾病的发病机制,包括经典的心身疾病、躯体化障碍、疑病症和躯体形式疼痛障碍。不幸的是,评估述情障碍结构的研究在进行时很少关注评估用于测量它的各种量表的心理测量特性。在两项独立研究中,我们考察了各种量表和项目特性,以及明尼苏达多相人格调查表述情障碍量表(MMPI-A)的因子结构和效度,MMPI-A是评估述情障碍最常用的量表之一。在研究1中,构成MMPI-A的22个项目是从一个计算机化的MMPI数据库中提取的,该数据库包括精神科住院患者和门诊患者的独立样本。两个样本的项目与量表的特性均较差,内部信度仅处于中等水平。因子分析得出的因子与述情障碍结构的理论领域相关性较差。在研究2中,我们发现该量表的效度几乎得不到支持,因为根据MMPI-A被认定为述情障碍和非述情障碍的患者在几个理论相关量表上并无差异。这些结果严重质疑了MMPI-A在研究述情障碍结构方面的价值。