The relationship between the writing competence and reading competence of Hong Kong fourth-grade students in the progress in international reading literacy study = 香港小四學生在「全球學生閱讀能力進展研究(PIRLS)」的書寫表達能力與閱讀理解能力的關係

Abstract

The study investigated students’ competence in producing written Chinese during reading comprehension tests, and its relations with Chinese reading comprehension performance of Grade Four students in Hong Kong. The study examined the test booklets completed by primary school students in Hong Kong as part of their contribution to the “Progress in International Reading Literacy Study (PIRLS)”. From the 4,712 booklets of PIRLS 2006 databank, 325 booklets were randomly extracted for the present study. In the first part of the study, a comparison was made between students’ responses to multiple choices (MC) and to open-ended (OE) items. An item effect was only statistically evident in responses to questions that “examine and evaluate content, language, and textual elements”. Noteworthy individual differences were found in the students’ responses to multiple choice questions that “interpret and integrate ideas and information.” The findings suggest that items pertaining to use of higher level cognitive processing helped distinguish advanced readers from their less able peers. The investigation of fourth grade students’ writing competence during reading comprehension was then conducted. Measurement procedures were used to assess the accuracy and pertinence of the students’ written answers. The students’ writing ability were derived from the length and depth of the answers, type of sentence utilized, sentence complexity, wording, paraphrasing and idea unit estimates. Content analysis, descriptive analysis, correlational analysis and multi-variant statistical procedures were applied to investigate the content of the writing produced by student participants and its relationship with the student body’s outcome of reading tasks. Regarding students’ writing ability evident in responses on the reading comprehension tests, analyses revealed that single sentences (simple or complex) were mostly used in students’ responses to open-ended questions. In terms of lexical expression, it was found that students tended directly to copy vocabulary, phrases or clauses from the passages before them, probably in an effort to avoid errors of expression. In other words, short answers with exact wording from the texts were usually presented, possibly reflecting a sizeable group of students who had passive attitudes towards reading comprehension tasks. Multiple regression analyses suggested that correct responses to open-ended questions could be predicted on the basis of students’ ability on paraphrasing, sentence construction as well as length of answers. This study contributes to the investigation of reading comprehension assessment in various ways. Firstly, the study is the first major attempt in Hong Kong to examine primary school students’ Chinese writing ability estimated and derived from their reading comprehension performance, as evident in their written test responses. Secondly, the study analyzed students’ writing ability during their performance on high-quality reading comprehension tests and identifies cognitive and linguistic features that help predict students’ performance on taxing reading tasks and their written responses. Furthermore, the study also informs the design of reading assessment measures, impactful reading instruction and the design and form of reading comprehension practice text books. It also suggests future research into the promotion of students’ reading and writing capability.

本研究旨在探討香港小四學生在中文閱讀理解測試的寫作能力，與閱讀理解能力的關係。本研究的目的，是要了解評估學生閱讀能力的合適題型和組合，以及補足現時閱讀能力評估研究的不足。研究尤其聚焦於評估學生在閱讀理解測試時的中文書寫作能力，分析學生的答題的中文書寫表達能力的構成變項，再跟他們在PIRLS 2006測試閱讀成績掛勾，驗證兩者的關係。 2006 年，全球學生閱讀能力進展研究（PIRLS）調查了香港4,712名香港小四學生的中文閱讀能力，所得的研究資料成為了PIRLS 2006數據庫。本研究從PIRLS 2006數據庫隨機抽樣選出325名學生的閱讀理解測驗本，進行文本、數據分析和相關分析，以了解小四學生中文閱讀理解的答題能力，包括多項選擇題（MC）和開放式答題（OE）的答題能力。 研究結果顯示，香港小四學生作答中文閱讀理解的MC和OE表現相近，兩種題型的差異對學生閱讀能力的影響，出現在作答第四層「評價篇章內容及語言形式」的題目。此外，學生「第三層：整合並解釋篇章」MC的表現差異較大，證明這類題目是分辨學生能否掌握高階閱讀能力的「試金石」，是甄別具有良好閱讀能力學生的重要工具。 通過相關分析篩選部分變項後，研究員運用迴歸分析驗證學生的OE問題書寫表達能力變項，對OE題目答對率的預測力。結果顯示「改述」、「句法」和「答案長度」三個變項對OE題目答對率的具有一定預測力，其中具有最強預測力的變項是「改述」。 本研究的貢獻在於運用了規模龐大的PIRLS資料庫，和小學生真實的閱讀理解答案作為研究材料，對於中文閱讀理解評估的研究具有重要的意義；同時，本研究建構了學生進行閱讀理解時書寫表達能力的理論與分析框架；在構成書寫表達能力的項目上，結合改述能力的理論和香港學生作答OE題目的語言特點，制定了改述能力的評級。研究員提出的一套研究設計和方法，更為將來在中文閱讀評估領域提供了一個參考。