目的：本研究之目的在瞭解團體治療過程中，中文版團體氣氛量表之信度及效度。
方法：本研究運用門診憂鬱症動力性人際關係團體心理治療資料，檢定中文版團體氣氛量表之內部信度及建構效度。
結果：中文版團體氣氛量表之投入、衝突及逃避之內部一致性檢定Cronbach alpha 係數分別為0.89、0.83 及0.20；折半信度（split-half reliability）分別為0.81、0.83 及0.06。經由因素分析可以發現量表包含投入、衝突及逃避等三個主要因素，但逃避的題項則與原始分類有較大的差異。
結論：中文版團體氣氛量表透過本研究的信效度檢定，發現投入及衝突有良好的信度及效度，可以廣泛運用在團體治療之評估上，逃避的題項則需調整。

Object: The purpose of this research was to validate the Chinese version Group Climate Questionnaire.
Method: We tested the internal reliability and construct validity by analyzing the data from dynamic interpersonal group psychotherapy in the outpatient clinic.
Result: The internal consistency Cronbach alpha coefficients of engagement, conflict andavoidance were 0.89, 0.83 and 0.20. The Guttman Split-Half Coefficients of split-halfreliability were 0.81, 0.83 and 0.06 for engagement, conflict and avoidance, respectively.By factor analysis, we found 3 components including engagement, conflict andavoidance. Items of engagement and conflict were compatible with the original Englishversion but items of avoidance were not.
Conclusion: The subscales of Chinese version Group Climate Questionnaire includingengagement and conflict have good reliability and validity. It is appropriate to apply these2 subscales in the evaluation of groups. The avoidant subscale needs some adjustment forfurther clinical application.