中文版多向度完美主義量表之心理計量分析

Psychometric Analyses of the Chinese Version of the Multidimensional Perfectionism Scale (CMPS): Reliability and Validity

研究目的：本研究旨在修訂中文版多向度完美主義量表，並進行信效度分析；也驗證Frost的六向度完美主義建構與二因素次層結構。根據文獻分析，本研究假設「過度在乎錯誤」、「對行動的懷疑」、「父母批評」及「父母期望」分量表分數，與強迫症狀或憂鬱症狀量尺分數呈低度相關；「個人標準」分量表分數與憂鬱症狀分數及強迫症狀量尺分數呈正相關；而「組織性」分量表分數與憂鬱症狀或強迫症狀分數則無關；並據此進行量表之效標效度分析。研究方法：透過分層叢集取樣，由南區大專校院大學生抽樣有效樣本2,190名；自填中文版多向度完美主義量表、身心健康評估量表。研究結果：全量表內部一致性為.938，分量表內部一致性介於.736~.903，六向度建構及二次層因素建構適配度指標良好，量表與各效標間關係亦支持上述假設。研究結論：中文版多向度完美主義量表具良好之信度、效標關聯效度，並支持Frost的六向度完美主義建構與二因素次層結構模型。

Purpose: The goals of this study were to examine the reliability and validity of the Chinese version of the Multidimensional Perfectionism Scale (CMPS), to examine the sixdimensional model of perfectionism with two second-order factors as proposed by Frost, and to explore the relationships among the six dimensions, two second-order factors, depressive symptoms, and obsessive-compulsive symptoms. Based on a literature review, the following hypotheses were examined: the subscale scores on “Concern over Mistakes”, “Doubting of Action”, “Parental Expectation”, and “Parental Criticism” had positive relationships with both the depressive and obsessive-compulsive symptom scores; the “Personal Standard” subscale score had a weak relationship with the depressive and obsessivecompulsive symptoms scores; the “Organization” subscale score was not correlated with either depressive or obsessive-compulsive symptom scores. Methods: Using stratified and cluster sampling, a total of 2,190 college students (mean age = 19.718) were recruited from southern Taiwan and administered the CMPS, Ko Depression Inventory, and Symptoms Check List-90-R. Results: The Cronbach’s alpha coefficient was .938 for the CMPS and ranged within between .736 and .903 for the subscales. Moreover, the six-dimensional model of perfectionism with two-second order factors had good model-fit evaluations. In addition, the “Concern over Mistakes”, “Doubting of Action”, “Parental Expectation”, and “Parental Criticism” subscales had positive relationships with the depressive symptom and obsessive-compulsive symptom scores. The “Personal Standard” subscale score had weak relationships with the depressive and obsessive-compulsive symptoms scores. The “Organization” subscale score was not correlated with either depressive or obsessivecompulsive symptom scores. Results supported all hypotheses. We also found that males had higher scores than females on “Parental Expectation” and “Doubting of Action”. Conclusions: The CMPS is a reliable and valid measurement of perfectionism. Moreover, Frost’s six-dimensional model of perfectionism with two second-order structures was supported.