| Comprehension monitoring strategies employed by reading-disabled (n = 25) and non-disabled (n = 35) adolescents were compared as they read a series of 20 independent sentences containing either a syntactic or semantic error. Purposes included comparing the two levels of reader on error detection rates, specific error recovery strategies employed, reading component skill variances, basic eye movement patterns (e.g., fixation times), and comprehension theory utilization. The three independent variables included reading level (disabled/non-disabled), two levels of syntactic error (ambiguous/nonambiguous), and two levels of semantic error (anomalous/nonanomalous). Dependent measures included eight eye movement measures, a set of comprehension questions, scores from six tests assessing component skills of reading (decoding, attention, linguistic competencies, STM, sequencing abilities, and spatial skills), as well as a qualitative analysis of eye movement patterns. This 2 x 2 x 2 repeated measures, three-factor mixed design used MANOVA, MANCOVA, ANOVA, multiple regression, Chi Square and Tukey's HSD in data analysis.;Results indicated greater similarity among the two reading groups in error detection and recovery from syntactic and semantic inconsistencies than had been previously reported. Both groups showed equivalent and high levels of error detection under low decoding demands. Seven newly-defined error recovery strategies were consistently exhibited by both disabled and non-disabled readers in similar usage frequencies. Few individual differences were noted. The most frequently employed error recovery patterns by both groups appeared strategic and economical; there was little evidence of disorganized eye movement patterns among either group. Discrepant results between retrospective reports and eye movement measures of detection were found, questioning the validity of the former. Component skill analyses indicated significant group differences, with sequencing ability being an important predictor of both eye movements and comprehension among the two groups. The Garden-Path Theory of Comprehension was partially supported in both groups, although methodological problems in theory support were noted. |
