An Investigation Of Factors Affecting Reading Test-taking Strategy Use

A review of the literature shows that both individual and task characteristics may affect the respondents' strategy use in taking reading comprehension tests. The current study attempts to investigate whether task characteristics including item types, item difficulty and text types affect the reading test-taking strategy use by the Chinese senior high school students who sit for the National Matriculation English Test (NMET). In order to accomplish this purpose, a sample of fifty-five students was chosen as subjects in the study.Both qualitative and quantitative analyses were used in the study. A multiple-choice reading comprehension test was administered as research task. The test includes three item types respectively reflecting the students' understanding of the main idea of the texts, the ability to draw simple inference, and understanding of the detailed information in the texts. Two kinds of text types were used in the test: Practical text;and Nonpractical or General text. Nevo's (1989) checklist of test-taking strategies for answering multiple-choice reading comprehension questions was used as the instrument to elicit the students' strategy use during the test taking. Item difficulty values were computed with reference to the Classical True Score model. Chi-square analysis was used to compare the frequencies of strategy use across different levels of the variables of interest.When the experiment was completed, only 21 of the students were found to have completed all the tasks. Results show that standard deviation of their strategy use frequencies is 98.12, suggesting an unbalanced use of the strategies. Four out of the fifteen strategies provided in the checklist have been left unused. Only fourstrategies have been used with frequencies above the mean (about 77). Seven of the reported strategies were used with frequencies below the mean, six of which had occurrences only ranging from 1-14. Significant results were found only in the use of the four most frequent strategies: (1) Background Knowledge is the only one that has been affected by all the three variables;(2) Process of Elimination was influenced by both item types and item difficulty but not by text types;(3) While the strategy Returning to the Passage was affected by item difficulty, not by item types or text types, the strategy Clues in the Text was influenced by item types, not by item difficulty or text types.This study may have implications for the construction of NMET reading comprehension test and the reading instruction in the high school classroom as well.