本研究主要目的在修訂「感恩量表」（Gratitude Questionnaire, GQ）以建立適用於臺灣地區之中文版。本研究將 McCullough、Emmons和Tsang（2002）所發展之感恩量表翻譯修訂為中文，以大一至大四學生為對象，先採立意抽樣選取 214人預試，進行項目分析、探索性因素及信度分析，並考量原始 GQ因素結構，建立三個中文版的競爭模式。續以非隨機分層抽樣選取 750人正式施測，做為建模及驗證兩組樣本，進行驗證性因素及效標關聯效度分析。本研究結果發現，由 5題所組成的單因素 GQ中文版具有合理適配、良好效標關聯效度、合理的組合信度與變異數平均解釋量。此外，本研究發現女大學生較男大學生有更高程度的感恩意向；有宗教信仰的大學生也較無宗教信仰之大學生有更高程度的感恩意向；但不同年級的參與者在感恩意向的表現上則無差異。 The main purpose of this study was to establish Chinese version of the Gratitude Questionnaire (GQ) in Taiwan. The researcher first translated the GQ into a Chinese version and conducted it to freshmen to senior students in universities as participants through two stages and a series of analytic procedures. The researcher first used 214 participants by using purposive sampling in the pretest to conduct item analysis, exploratory factor analysis, reliability analysis and took ordinary factor structure of the GQ into consideration, then proposed three alternative-competitive models (six items for two factors, five items for one factor, six items for one factor). The researcher then used another 750 participants by using non-random stratified sampling to conduct confirmatory factor analysis and criterion-related validity, including one group for calibration sample and the other group for validation sample. The results suggested that the Chinese version of GQ composed of five items not only has good goodness-of-fit and criterion-related validity, but also has good composite reliability and average variance extracted. The Chinese version of GQ therefore has good reliability and validity. This study also found that the girls outperformed the boys on the gratitude; moreover, the students who had religious belief outperformed the ones who had no religious belief on the gratitude, but no significant differences were found among the four grade levels.
教育與心理研究, 35(4),103-134 Journal of Education & Psychology