二年級「持續型」與「晚發型」識字困難學童早期區辨效能之檢測

Predicting Persistent or Late-Emerging Difficulties in Reading Chinese Characters in Grade-2 Children

Purpose: Recent studies have proposed that testing for reading-related cognitive skills can identify children with difficulties in reading Chinese characters in Grade 1. Research has also suggested that some children demonstrate adequate reading achievement when they start learning to read but lagmarkedly behind their peers in subsequent grades. These children are referred to as late-emerging poor readers.This study was intended to (1) investigate the importance of a battery of reading-related skills, including phonological awareness in syllable and tone, Tzu-yin decoding, visual memory, and rapid automatized naming (RAN) in character reading from preschool to Grade 2; (2) determine the differences in reading-related skills among groups of children with or without persistent or  late-emerging character-reading difficulties; and (3) examine the accuracy of predicting persistent or late-emerging difficulties in reading Chinese characters in children. Methods: This was a 3-year longitudinal study. Although 330 preschoolers were recruited, only 230 remained in the cohort at the end of the study. The participants reading-related skills were assessed in the last semester of preschool (April to June 2008). Their progress was observed and their reading ability measured at 3 time points: the beginning of Grade 1, the end of Grade 1, and the end of Grade 2. Fifteen of the 230 students had difficultyreading characters in both Grades 1 and 2, and were deemed as having persistent difficulties. Fourteenchildren who started having difficulties in Grade 2 were defined as children with late-emerging difficulties. Thirty-five children whose z score of character-reading ranged from 0to 0.5 in both grades were classified as the control group. Correlation, analysis of variances, and discriminant function analysis were employed to analyze the data. Results/ Findings: The results revealed that Tzu-yin decoding and RAN in preschool were relevant to character-reading skills in Grades 1 and 2. The importance of phonological awareness gradually increased with experience in reading, whereas visual memory was substantially associated with character reading only at the initial stage of learning to read. The analysis of variances suggested that Tzu-yin decoding and RAN differed significantly among the groups. Children with persistent difficulties in character-reading scored significantly lower than normal readers in Tzu-yin decoding and RAN. Late-emerging poor readers scored higher thanpersistent poor readers and lower than normal readers in Tzuyin decoding and RAN; however, the differences between late-emerging poor readers and normal readers were not statistically significant. The results of discriminant function analysis provided one crucial discriminant function thatrecognized early characterreading, Tzu-yin decoding, and RAN as significant identifiers of reading difficulties. The classification ratewas 71.9% accurate. Early character-reading was shown to be the most accurate identifier, and offered no false negative values. Conclusions/Implications: The most significant early predictors of reading difficulties in Grade-2 children were Tzu-yin decoding, RAN, and early character-reading. Persistent poor readers might have more difficulties in learning to read than late-emerging poor readers do, and the discriminant function of this study accurately identified persistent poor readers. Althoughthe discriminant functionhad limited success in identifying late-emerging poor readers;the discriminant function predicted 100% of children with persistent difficulties in character reading and 80% of those with late-emerging difficulties from preschool to Grade 2, when the data was analyzed by combining the persistent and late-emerging groups.