九十八年學測閱讀測驗考生作答策略之初探

Strategies in Response to the Reading Comprehension Items on the 2009 General Scholastic Ability English Test

鄭瑞芝(Jui-chih Cheng)

指導教授 ： 張武昌

國立臺灣師範大學/文學院/英語學系/博士(2012年)

摘要

本研究探討高三考生在2009 年學科能力測驗閱讀測驗部分的作答過程與
策略運用。分別針對考生在閱讀測驗部分普遍的策略使用、不同題型的策略使
用、以及難易度不同考題的策略使用進行分析與探討。2009 年學科能力測驗閱讀測驗部分共有四篇文章，每篇文章長度約248 至298 字。十八位高三考生(高、中、低程度考生各六位)於2009 年學科能力測驗結束後兩天之內參與本研究。在資料採集的過程中，十八位高三考生針對同樣一份考題的閱讀測驗部分進行作答，並詳細口述其閱讀過程、每題作答過程與選項依據。研究者參酌並修正Cohen與Upton 於2006 年所作的閱讀及考題作答策略分類表，將十八位高三考生的口述語料進行閱讀及考題作答策略分類，並做進一步的分析與探討。
研究發現十八位高三考生在2009 年學科能力測驗閱讀測驗十六道考題的
作答過程中，共運用十八項閱讀策略與三十六項考題作答策略。其中高成就考生在作答時，逾八成的考題會依據文章內容做選項的選取或刪除；中等程度考生在作答時，近五成的考題會依據文章內容做選項的選取或刪除；低成就考生只有三成的考題會依據文章內容做選項的選取或刪除。
回答閱讀測驗十六道題目時，多數高成就考生針對不同題型的考題，分別
所使用的閱讀與作答策略相當一致。中、低成就考生則使用不同的作答策略。
回答文章主旨或用意的題型時，多數高成就考生會直接依據文章的大意做
選項的選取或刪除。在回答文章細節、字義、指涉等題型時，多數高成就考會
詳讀文章一部分以搜尋線索，再依據文意做選項的選取或刪除，多數高成就考生鮮少運用字面作答技巧。
回答2009年學測閱讀測驗難度最高的考題(第44題)時，考生需要統整文章
段落之間的訊息，再針對四個選項作真偽的判別。參與本研究的十八位高三考生中，多數考生在閱讀題目之後會再詳讀文章一部分搜尋線索，然而只有高成就考生實際依據文意做選項的選取或刪除，中、低成就考生皆未依據文意做選項的選取或刪除。中、低成就考生無法理解選文內容與選項字義，因而使用較多的作答策略，其中又以運用先備知識與字面作答技巧居多。
回答2009 年學測閱讀測驗難度極低的一道考題(第48 題)時，考生需要在
閱讀題目之後，詳讀文章一部分以搜尋線索，再依據文意判定某一名詞的指涉。
參與本研究的考生相對使用少量的應答策略，且十八位考生全部答題正確。然而中、低成就考生在依據文意判定指涉的策略使用上，遠低於運用字面作答技巧的策略使用。五成程度中等的考生與全數低成就考生皆未依據文意做名詞的指涉，而直接運用字面作答技巧，選取出現文章關鍵字的選項。從本研究高三考生的應答策略中，發現此道考題的設計似無法有效評量考生對名詞指涉的能力。
本研究發現程度不同的考生在2009 年學測閱讀測驗文章的閱讀與作答的
過程不甚相同。高成就考生已具備該測驗所欲測試考生的英文能力，因而能夠有效率地閱讀文章並成功回答多數的題目。低成就考生尚未具備一定程度的英文能力，無法理解選文內容與多數選項字義，因而使用較多字面作答技巧。
本研究結果可提供國內英語教師與大考中心作為日後命題與審題的參酌。
在設計考題時，正答選項中出現文章關鍵字的考題，似比正答選項中未出現文章關鍵字的考題簡單；誘答選項中出現文章關鍵字卻陳述不真事實的考題，似比誘答選項未出現文章關鍵字的考題更具挑戰性。在設計考題時，應避免正答或誘答選項能以考生先備知識判定真偽的題目，並宜注意考題難度的排序，盡量將難度低的考題先置於難度高的考題，以增加考生答題的成就感並提升閱讀動機。
本研究結果亦提供英語教師作為教學上的參考。英語教師可以加強學生基
本語言能力與善用閱讀策略，協助中、低成就學生克服閱讀的困難，英語教師並宜挑選多種適合學生程度的文章讓學生廣泛閱讀，從閱讀中提升閱讀能力。

關鍵字

閱讀測驗 ； 作答策略 ； 閱讀策略 ； 考題難度 ； 高中低成就考生 ； 命題效度 ； 外語評量 ； 學科能力測驗 ； 中成就考生 ； 低成就考生

並列摘要

This study describes the responding strategies that test takers used on the reading comprehension subtest of the 2009 General Scholastic Ability English Test [GSAET]. The investigation focused on the examinees’ response strategies for: (1) 16 reading comprehension items in general; (2) 7 types of reading comprehension items; and (3) the most and least challenging test items. Verbal report data were collected from 18 12th graders across proficiency levels, i.e., 6 high achievers, 6 average achievers, and 6 low achievers. Participants worked on the reading comprehension subtest of the 2009 GSAET, containing four 248-298 word passages with four items. Participants’ verbal report was evaluated to determine strategy use based on the modified versions of Cohen and Upton’s reading and test-taking strategies coding rubrics (2006).
The participants used a total of 18 reading strategies and 36 test taking strategies in response to the 16 reading comprehension items. The high achievers selected or discarded the options through textual meaning 80% of the time; the average achievers, 50% of the time; and the low achievers, only 30% of the time.
In response to the reading comprehension items, the high achievers generally showed a consistent pattern in the use of the main response strategies for different types of questions, whereas the average and low achievers resorted to a variety of test taking strategies.
In response to global questions, e.g., questions on the main idea or purpose of the passage, the high achievers generally selected the options through passage overall meaning. In response to local questions, e.g., questions asking for vocabulary meaning, a specific referent, inference, specific information, cause-effect relationship, or details of the passage, the high achievers generally went back to the passage, read carefully for the clues, and selected the options through vocabulary, sentence, or paragraph meaning. They generally selected or discarded the options through textual meaning instead of test-wiseness strategies.
In response to the most challenging question, Q44, which required the respondents to integrate information conveyed across paragraphs and justify the correctness of the statements in four options, most of the participants read the question and then went back to the passage, carefully reading a potion of the passage to look for clues. While the high achievers generally selected and discarded options through textual meaning, none of the average and low achievers selected the option through textual meaning. Showing difficulty comprehending the passage and wrestling with the option meaning, the average and low achievers used more test-taking strategies in response to the question. They relied more on their background knowledge and the key word association strategy in their option selection.
In response to one of the least challenging items, i.e., Q48, all of the participants manipulated relatively fewer response strategies and successfully selected the option. But the strategy of verifying the referent was used at a much lower frequency rate than the test-wiseness strategy of key-word matching among the average and low achievers. Half of the average achievers and all of the low achievers selected the option through key-word matching strategy. The response strategies thus did not seem appropriate for the purpose of the item and provided weak evidence for theory-based validity.
This study showed that examinees of different proficiency levels processed the passages/tasks differently. The high achievers, whose English proficiency had reached a certain level required of the 2009 GSAET, were able to read the passages efficiently and completed most of the test items successfully. The low achievers, whose English proficiency had not reached a certain level to cope with most of the test items, wrestled with the meanings of the words in the passages/tasks and failed to process the passages/tasks globally.
The findings provide insights into the construction of L2 reading tests. They suggest that questions with the correct option containing key words in the passage are likely to be easier than questions without; questions with distracters containing words in the passage but describing something irrelevant to the passage are likely to be more challenging than those without; and questions with options involving statements which can be judged wrong from the examinees’ background knowledge do not make attractive distracters. They also confirm the importance of sequencing test items by difficulty, with easy items preceding challenging ones.
The findings also provide pedagogical implications, suggesting that L2 teachers may assist learners to cope with difficulties in reading by improving word-level competences and promoting use of comprehension strategies from a range of texts appropriate to learners’ proficiency levels.

並列關鍵字

reading comprehension ； test-taking strategy ； reading strategy ； item difficulty ； high achiever ； validity ； second language assessment ； GSAET ； average achiever ； low achiever

參考文獻

殷允美、楊泰雄、葉錫南、林秀慧、游春琪(2008)。95課綱試題研究工作計畫--英文科研究報告。台北：大學入學考試中心。[Yin, Y. M., Yang, T. H., Yeh, H. N., Lin, H. H., & Yo, C. C. (2008). Project report on item development based on the 2006 Senior High School Curriculum Guidelines for the English Subject. Taipei: College Entrance Examination Center.]

strategies on English language learners’ test performance. Educational Measurement: Issues and Practice, 19, 16-26.

Afflerbach, P. (2007). Understanding and using reading assessment K-12. Newark,

DE: International Reading Association.

Alderson, J. C. (1984). Reading in a foreign language: a reading problem or a