「閱讀理解篩選測驗」的再驗證和應用

Validation and Application of the Reading Comprehension Screen Test

本研究以Rasch 分析重新檢驗「國民小學二年級閱讀理解測驗（A）」（柯華 葳、詹益綾，2007）的信、效度，並探討其應用結果。研究對象來自北、中、南、 東四區隨機選取的882 名國小二年級學童。此閱讀理解篩選測驗分為句子理解7 題、命題組合5題、短文理解7題，共19題單選題。結果顯示，符合命題組合、句 子理解與短文理解三向度結構模式，亦即模式與資料間有良好的吻合度，而且並無 性別上的差異試題功能（DIF），但在句子理解和命題組合向度各有一題具有地區 性的DIF。該測驗的題目難度偏低且分配較集中，受試者能力明顯高於試題難度。 進一步分析測驗應用結果發現，有15.1%的受試學童可能有閱讀理解困難。本研究 亦針對測驗的應用提出建議。

This study aimed to validate “the Reading Comprehension Test for Second Graders (A)” (Ko & Chan， 2007) using Rasch analysis and to explore the application results. A total of 882 second graders from four geographic areas in Taiwan completed this test. The Reading Comprehension Screen Test consisted of 19 multiple-choice questions including 7 questions on sentence comprehension， 5 questions on naming task， and 7 questions on passage comprehension. The results revealed that the three dimensions of the Reading Comprehension Screen Test showed good model-data fit， and no gender DIF was found. However， regional DIF was shown respectively in one question on the sentence comprehension and on the naming task. The data manifested that the difficulty of the test questions was relatively low and skew as the ability of the participants was remarkably higher than the level of the test questions. There were 15.1% of participants at risk of reading comprehension difficulty. Furthermore， concerns related to the use of the test were discussed.