사실적 이해 능력 평가를 위한 한국어능력시험 읽기 문항의 환언 적절성 연구

Abstract

The purpose of this research is to set appropriate standards for paraphrasing when writing multiple-choice reading questions for testing literal comprehension ability and to make suggestions for appropriate paraphrasing based on the results obtained through test-item analysis. Three main criteria were set for evaluating the level of adequacy in paraphrasing, which includes the adequacy of the paraphrased unit, the consistency of meaning between the original and the paraphrased texts, and the linguistic level of paraphrased text. The unit of the paraphrasing should be done at the syntactic level rather than at the lexical level, and the meanings of the original and the paraphrased texts must exactly match. In addition, the language level of the paraphrased text ought to be equal to or lower than that of the original text. Based on these criteria, the items measuring literal comprehension ability in the TOPIK(Test of Proficiency in Korean) were analysed. First, observing the adequacy of the paraphrased unit, it was found that although paraphrasing was done mostly at the syntactic level, there were more lexical-level paraphrasing in the beginner level than the intermediate and advanced level. Next, the results showed that there were more discrepancies of meaning in the intermediate and advanced level than the beginner level, and more cases of inconsistencies in syntactic-level paraphrasing than thelexical-level paraphrasing. Lastly, it revealed that the ratio of the linguistic level of the paraphrased text being higher than that of the original text were somewhat higher in the beginner level than the intermediate and advanced level. Thus, when creating multiple-choice items for assessing literal comprehension ability, it is necessary to construct the questions based upon these criteria of the adequacy and to check the items against the criteria in the editing stage.