近年來，用來測試學生學術寫作能力的整合性試題越來越受到歡迎。有許多營利性及非營利性的英語測驗都包含了閱讀寫作及圖表寫作的試題。一般而言，傳統獨立寫作試題考核學生對於某一議題本身的思考能力，而整合性英文寫作試題更進一步考核學生是否能擷取、分類並統整參考資料中的重要資訊。相較於獨立寫作試題，整合性寫作試題更接近學術環境中的溝通時況，真實的模擬大學校園中可能遇到的寫作需求。因此整合性寫作試題常被視為可以增進考題的真實性及公平性。同時，也因為此種考試題型測試學生長期累積的英文讀、寫之實力，學生的英語學習也能透過此種評量方式而得到良好的影響，亦即帶動正面的回沖效應(backwash effects)。然而，整合性寫作測驗的整合特質也引發一些與建構效度 (construct validity) 相關的爭議。由於測驗的建構效度乃是測驗核心，是用來說明測驗成績 (test interpretation) 及成績使用 (test use)的基礎，深入的了解這兩類試題的構念 (construct) 是必要的。本研究以中山醫學大學學生為研究對象，採結構方程模組與質化研究之途徑來瞭解兩類整合性寫作試題 – 閱讀寫作及圖表寫作試題，及學生寫作策略使用是否影響其各面向的寫作表現。這個研究有助於了解台灣英語學習者如何擷取閱讀及圖表綱要來完成寫作考試。藉由追蹤及比較學生的寫作策略使用與各向度之寫作能力，此研究成果也能讓英文教師更了解寫作測驗的本質以提供有效的策略教學來提升大專學生英語寫作能力。
In recent years, the use of integrated writing tasks for assessing academic writing in English has been increasing in popularity. A number of commercial language tests and in-house writing tests have included integrated reading-based writing tasks and graph-based writing tasks in their assessment batteries. These tasks are proposed to offer more authenticity, improve fairness, and provide positive washback effects of the test on learning and teaching of English. However, the integrated nature of these tasks can pose issues of test development and constructed-related validity. Given that the inferences made from test scores depend upon the construct of the measure, it is important to have a deeper understanding of the constructs measured by these two task types. This project investigates the effects of the types of source texts, namely verbal and visual texts, in the prompts and test takers’ strategy use on their test performance. Approximately one hundred and fifty undergraduate students will be recruited to take reading-based writing and graph-based writing tests, followed by a strategy inventory on how they thought while completing the test. Twenty students, ten from the high achievement group and ten from the low achievement group, will be selected for retrospective interviews. The data will be analyzed quantitatively and qualitatively. Exploratory factor analysis (EFA) will be used to identify the link between strategy use (latent variables) and student essays (measured variables). Drawing upon previous writing research, confirmatory factor analysis (CFA) will be utilized to test the hypothetical relations between observed and latent variables. A structural equation modeling (SEM) will be used to model relationships between students’ self-reported writing strategy use and their writing performance. The data collected from retrospective interviews will be analyzed to provide supplementary information in interpreting the quantitative data. The results of the study will have implications for second language writing assessment and instruction.
