Mehrabian A
Department of Psychology, University of California, Los Angeles, USA.
Genet Soc Gen Psychol Monogr. 1995 Aug;121(3):339-61.
The rationale for a general three-dimensional approach to measuring emotions (the Pleasure-Arousal-Dominance, PAD, Emotional-State Model) is described. Evidence bearing on two early versions of the PAD scales is reviewed, and three studies report additional development and refinement of a final set of PAD scales. Factor analysis of 58 PAD items in Study 1, 41 PAD items in Study 2, and 34 PAD items in Study 3 consistently yielded three nearly orthogonal Pleasure, Arousal, and Dominance factors. The final 16-item State Pleasure, 9-item State Arousal, and 9-item State Dominance scales had alpha reliability coefficients of .97, .89, and .80, respectively. Across Studies 2 and 3, intercorrelations among the latter three scales did not exceed .09 in absolute value, showing that the scales provided a parsimonious base for the general assessment of emotional states. Discriminant and construct validity evidence for the scales were reviewed. Potential applications of the scales were noted.
本文描述了一种用于测量情绪的通用三维方法（愉悦-唤醒-支配，即PAD情绪状态模型）的基本原理。回顾了与PAD量表两个早期版本相关的证据，三项研究报告了最终版PAD量表的进一步开发和完善情况。研究1中对58个PAD项目、研究2中对41个PAD项目以及研究3中对34个PAD项目进行的因子分析，均一致得出了三个近乎正交的愉悦、唤醒和支配因子。最终的16项状态愉悦量表、9项状态唤醒量表和9项状态支配量表的α信度系数分别为0.97、0.89和0.80。在研究2和研究3中，后三个量表之间的相互关联绝对值均未超过0.09，这表明这些量表为情绪状态的总体评估提供了一个简洁的基础。文中还回顾了这些量表的区分效度和结构效度证据，并指出了量表的潜在应用。
