Abstract
PURPOSE: The purpose of this study was twofold. The first aim was to explore differences in profiles of past tense marking in oral reading of school-age children with specific language impairment (SLI). The second aim was to explore the potential of past tense marking in oral reading as a clinical marker of SLI in school-age children. METHOD: This study examined oral readings of connected text to describe the frequency and type of reading errors on regular and irregular past tense verbs for 21 children with SLI as compared to 30 children with typical language in Grades 2 and 3. Each past tense verb token was categorized into 1 of 6 mutually exclusive response types: (a) correctly marked past tense, (b) overmarked past tense, (c) bare stem, (d) other verb inflection, (e) nonverb, or (f) no response. Performance across groups was compared. Additionally, classification statistics were calculated at several cutoffs for regular past tense accuracy and regular past tense finiteness marking. RESULTS: For regular past tense, there was a significant group difference on accuracy. Children with SLI were less accurate at marking past tense when in oral reading than typical language peers; other response types did not differ. For irregular past tense, there were no group differences. In addition, there was a significant group difference on finiteness marking; this difference was driven by regular but not irregular verbs. A cutoff of 90% for regular past tense accuracy yielded moderate sensitivity and specificity; no cutoff for regular past tense finiteness marking yielded sensitivity above 70%. CONCLUSIONS: Regular past tense accuracy in oral reading provides promise as a clinical marker for diagnosing SLI in school-age children.