Language assessment with nonword repetition tasks (Schwob & Skoruppa, 2022)

Purpose: Over the last decades, many studies have documented the clinical potential of nonword repetition (NWR) tasks for detecting developmental language disorder in mono- (MON) and bilingual (BIL) children by unveiling their difficulties in short-term memory and phonological accuracy. However, the precise nature of the nonwords to be used and the best scoring methods remain under debate. Some authors (e.g., Gutiérrez-Clellen & Simon-Cereijido, 2010) support the use of “language-specific” nonwords designed for a given test language in standardized tests. Other authors (e.g., Chiat, 2015) advocate the use of “crosslinguistic” stimuli, thus allowing assessment independently of the languages spoken by the child.

Method: This research note compares two language-specific tasks (French vs. Portuguese) and a crosslinguistic NWR task in a population of 5- to 7-year-old MON and BIL children. Group comparisons (children with vs. without developmental language disorder; MON vs. BIL children), an error analysis, sensitivity and specificity calculations (assessed according to the recommendations of Plante and Vance, 1994, and Youden, 1950) are reported.

Results: All three tasks significantly differentiate children with and without developmental language disorder with large effect sizes but did not show an effect for bilingualism, which is encouraging for the BIL assessment. As expected, an influence of children’s age and length and complexity of the stimuli was also found. The language-specific French task was found to be the most sensitive (max. 88%) and specific (max. 92%); the crosslinguistic task also reached good accuracy percentages for the BIL group (max. 82% sensitivity and 84% specificity).

Conclusion: This research note adds to the evidence that NWR tasks are promising tolls for the identification of MON and BIL children with developmental language disorder.

Supplemental Material S1. Statistical results of the performances at the three NWRT, effect of clinical, language, and age status.

Supplemental Material S2. Scatter plots for the number of incorrect words and phonemes for the CL-NWR task.

Supplemental Material S3. Scatter plots for the number of incorrect words and phonemes for the French NWR task.

Supplemental Material S4. Scatter plots 477 for the number of incorrect words and phonemes for the Portuguese NWR task.

Supplemental Material S5. Statistical results of the performances at the three NWRT, effect of length, and complexity.

Supplemental Material S6. Results of errors in the three NWRT.

Supplemental Material S7. ROC curves for the different tasks and different scoring method.

Supplemental Material S8. ROC curves for the different tasks and different scoring method for monolingual and bilingual groups.

Supplemental Material S9. Items of the NWRT, length and complexity criteria.

Schwob, S., & Skoruppa, K. (2022). Detecting developmental language disorder in monolingual and bilingual children: Comparison of language-specific and crosslinguistic nonword repetition tasks in French and Portuguese. Journal of Speech, Language, and Hearing Research. Advance online publication. https://doi.org/10.1044/2021_JSLHR-21-00017