Edwards J R, Baglioni A J
Darden Graduate School of Business Administration, University of Virginia, Charlottesville 22906-6550.
J Appl Psychol. 1991 Apr;76(2):276-90. doi: 10.1037/0021-9010.76.2.276.
Although the Type A behavior pattern (TABP) is typically considered a set of distinct components, most studies of TABP have used global measures, which collapse several components into a single index. These measures are inherently multidimensional and, as such, contain several conceptual and methodological problems. In this study, data from 240 executives were used to compare global and component TABP measures as predictors of mental and physical symptoms. Global measures included the Bortner scale, the Framingham scale, and the Jenkins Activity Survey. Component measures were constructed by recombining items from the global measures on the basis of results from previous confirmatory factor analyses (Edwards, Baglioni, & Cooper, 1990). Results indicate that the component measures were superior to the global measures in terms of number of relationships detected, interpretability, and total explanatory power. Implications for research and practice are discussed.
尽管A型行为模式(TABP)通常被认为是一组不同的组成部分,但大多数关于TABP的研究都使用了综合测量方法,即将几个组成部分合并为一个单一指标。这些测量方法本质上是多维度的,因此存在一些概念和方法上的问题。在本研究中,使用了来自240名高管的数据,以比较作为心理和身体症状预测指标的综合和组成部分的TABP测量方法。综合测量方法包括博特纳量表、弗雷明汉量表和詹金斯活动调查。组成部分测量方法是根据先前验证性因素分析的结果(爱德华兹、巴廖尼和库珀,1990年),通过重新组合综合测量方法中的项目构建而成的。结果表明,在检测到的关系数量、可解释性和总解释力方面,组成部分测量方法优于综合测量方法。文中讨论了对研究和实践的启示。