本研究之目的在於探討中文版「情境─特質焦慮量表」之多元因素效度，以作 為衡鑑情境與特質焦慮的測量理論基礎。本研究以792 位高屏兩縣市的高中職二年 級學生為研究對象，結果顯示在一階層級的因素結構下，不論從傳統或現代測驗理 論，兩個量表都屬於一個多向度的因素結構，分別由情境、特質焦慮─正、反面措 詞等二個因素所組成，這二個因素彼此間具良好的區辨效度，且各題項間具有很好 的聚合效度；而在二階層級的因素結構下，兩個量表又都屬於一個單向度的高階因 素結構，分別代表情境、特質焦慮因素。另外，在一階層級的因素結構下，「情境 焦慮量表」或「特質焦慮量表」二個因素隸屬的各題項，都有三組題項因措詞相 似，而造成題項測量誤差彼此有關聯；「特質焦慮量表」有一個題項因措詞模稜兩 可，而有一個題項因測量不同概念導致橫跨兩個因素。等級反應模型顯示，兩個量 表各題項均具有很好鑑別力。因此，除「特質焦慮量表」的二個題項外，兩個量表 都具有很好的多元因素效度，由這些題項加總形成的分數，可以有效的反映這兩種 焦慮抽象概念，並且背後有良好的心理計量實證支持。

The purpose of this study was to examine the multidimensional factorial validity for the Chinese version of State-Trait Anxiety Inventory (C-STAI). Seven hundred and ninety two high school students from Kaohsiung and Pingtung Counties of southern Taiwan were invited to fill out the consent and this inventory. The findings show that each of those two scales belongs to a multidimensional factor structure which consists of two factors, state or trait anxiety - positive and negative wording in terms of the first-order level of factor structure, and there are good discriminant and convergent validity for each of those two scales. Each of those two scales is a one-dimensional second-order factor structure which represents the state or trait anxiety in terms of the second-order level of factor structure. Furthermore, there are moderate or high correlations between three pairs of the measurement errors since the wordings of those items may be very similar for those two scales. In addition, two items had cross factor loadings since one may be ambiguous wording and the other may measure different constructs for the Trait Anxiety Scale. Ignoring the controversial items above, the graded response model indicates that the remaining items have a good discrimination for each of those two scales. Therefore, the C-STAI had good multidimensional factorial validity except two items of ATrait.