초등 3-5학년 언어장애아동의 읽기이해모니터링

Abstract

The purpose of this study was to investigate reading comprehension monitoring including error detection and correction with written stories in school-aged children with specific language impairment. Thirteen children with specific language impairment (SLI) between Grade 3 and 5 and thirteen typically developing children matched on grade participated. All children with SLI obtained scores lower than -1.25 standard deviation of the normative mean of the Language scales for School-aged children which is the standardized language test in Korea (LSSC; Lee, Heo, & Jang, 2015) and scored 80 or higher on the Korean Comprehensive Test of Nonverbal Intelligence 2 (KCTONI-2; Park, 2014). Typically developing children obtained scores 80 or higher on the LSSC and KCTONI-2(Park, 2014). All of children with SLI and typically developing children have no history of known hearing, visual, neurological, emotional or behavioral problems. Three narrative stories were created for the reading comprehension monitoring. Because familiarity would play important role in reading comprehension, all stories were created by investigators. All three stories have 27 sentences and contained grade-appropriate vocabularies. In order to assess comprehension monitoring, stories contained three different types of errors. The errors which embedded within stories consisted of nonsense words (i. e. lexical inconsistency), violation of prior knowledge (i. e. external inconsistency) and internal inconsistency within the text (Baker, 1984; Markman & Gorin, 1981). The number of each error type was 6, so total number of error was 18. Comprehension monitoring tasks were presented in the following order: error evaluation and regulation. in the error detection task of comprehension monitoring, children were provided with written stories, asked to read stories, and detect when something is wrong. After completing the error detection task, they were introduced same written stories but errors were underlined. Children were asked to read stories again and correct underlined errors to make sense stories.

The following are the specific research questions : (1) Are there differences on performance of error detection and correction between children with SLI and typically children? (2) Are there differences on three types of error of the detection task between two groups? (3) Are there differences on three types of error of the correction task between two groups?

The results and discussions were as follows: First, children with SLI obtained mean scores of 4.39 on the error evaluation of the comprehension monitoring task while typically developing children achieved an average score of 11.85. Children with SLI obtained mean scores of 10.15 on the error regulation of the comprehension monitoring task while typically developing children have an average score of 14.46. There are significant differences on error detection and correction between two groups. The results showed children with SLI did demonstrate reading comprehension monitoring problems when compared to age-matched typically developing children. Specifically, children with SLI were more likely to have difficulty to detect problem in passages than to regulation problems in passages. The reason why children with SLI did show difficulty on error detection is that error detection is likely to related to high level of meta-cogntive process. Generally, error detection phase is related to metacognitive process while the error correction phase is related to general cognitive process (Pereia-Laird, 1997). Second, on the error detection task, children with SLI showed difficulties in all types of errors than typically developing children. Both groups are most likely to detect external inconsistency errors than other errors. Both groups showed most difficulty to find internal inconsistency errors than others. The reason for this is that in order to find internal inconsistency error, children need to storage previous information and process incoming information. Unlike typically developing children, children with SLI have difficulty to find even lexical inconsistency errors. Because children with SLI would be have deficits on vocabulary knowledge, they might have difficulty to find lexical inconsistency errors. Third, on the correction task, children with SLI showed difficulties in all types of errors than typically developing children. In addition, both groups are most likely to correct external inconsistency error type because external inconsistency errors were relatively clear semantic properties and comparatively easy. As the error correction task, children showed most difficulty to correct internal inconsistency. In summary, children with SLI did show difficulty of reading comprehension monitoring problems when compared to age-matched typically developing children and their poor reading comprehension monitoring ability could be negatively influenced on their reading comprehension performance. Their poor performance of monitoring tasks could be explained by lack of vocabulary knowledge and/or working memory. The results of this study suggest that monitoring intervention would be initiated in an effort to improve reading comprehension in children with SLI.