Objectives: The purpose of the current study was to examine the diagnostic accuracy of working memory (WM) tasks, including nonword repetition (NWR), sentence repetition (SR), and matrix tasks, as clinically efficient diagnostic tools depending on the scoring systems and task condition. Methods: Forty children with vocabulary delay (VD group) and 40 typically developing children (TD group) were administered a series of WM tasks (NWR, SR, and matrix tasks). Different scoring units were applied to each task (syllable vs. word for NWR task and word vs. sentence for SR task) while different conditions (forward matrix vs. backward matrix) were used for the matrix task. Separate discriminant function analyses were conducted with the scores calculated by the different scoring units or the different condition of the tasks as predictors. Results: Multiple discriminant function analyses for the NWR task yielded word units with higher overall classification values than syllable units, whereas for the SR task sentence units had a higher overall classification value than word units. For the matrix task, none of the conditions reached the level of significance in the discriminant function analyses. Conclusion: The results in this study support the diagnostic accuracy of the NWR task with word units and the SR task with sentence units in discriminating among the VD group and the TD group. However, caution should be used when regarding these matrix tasks as an accurate diagnostic tool, especially in young children.
Bibliographical notePublisher Copyright:
© 2017 Korean Academy of Speech-Language Pathology and Audiology.
- Diagnostic accuracy
- Discriminant function analyses
- Nonword repetition
- Sentence repetition