Shortened nonword repetition task

Purpose: The purpose of this research note was to validate a simplified version of the Dutch nonword repetition task (NWR; Rispens & Baker, 2012). The NWR was shortened and scoring was transformed to correct/ incorrect nonwords, resulting in the shortened NWR (NWR-S). Method: NWR-S and NWR performance were compared in the previously published data set of Rispens and Baker (2012; N = 88), who compared NWR performance in 5 participant groups: specific language impairment (SLI), reading impairment (RI), both SLI and RI, one control group matched on chronological age, and one control group matched on language age. Results: Analyses of variance showed that children with SLI + RI performed significantly worse than other participant groups in NWR-S, just as in NWR. Logistic regression analyses showed that both tasks can predict an SLI + RI outcome. NWR-S holds a sensitivity of 82.6% and a specificity of 95.4% in identifying children with SLI + RI. The sensitivity of the original NWR is 87.0% with a specificity of 87.7%. Conclusions: As the original NWR, the NWR-S comprising a subset of 22 nonwords scored with a simplified scoring system can identify children with combined SLI and RI while saving a significant amount of the needed assessment time. Supplemental Materials: