Comparing Specific Language Impairment and Hearing Impairment: Different Profiles in German Verbal Agreement Morphology

ABSTRACT

The study aims at identifying characteristic phenotypes for children with SLI and children with sensorineural hearing impairment (HI) in language and in domains associated with language. We focus on verbal agreement inflection and phonological short-term memory, phenomena that have been repeatedly found to be impaired in both groups of children. A nonword repetition task and an elicitation task on subject-verb agreement were conducted with three groups of monolingual German children: (i) 11 children with SLI, (ii) 10 children with HI (hearing loss between 38 and 75 dB) and hearing aids, and (iii) 10 typically developing children. Data analyses reveal quantitatively and qualitatively different performance patterns with respect to verbal agreement inflection in children with SLI and children with HI but no different patterns for nonword repetition. Our results not only identify deficits in verbal agreement inflection as a sensitive, specific, and selective clinical marker for SLI in German, but they also shed some light on the nature of the deficits that underlie SLI.

Acknowledgments

Our thanks go to the children and their parents for participating in this research and to the many practitioners and therapeutic institutions that helped in recruiting participants. We are grateful to Eva Wimmer, Johannes Hennies, Cornelia Nutsch, and Markus Tönjes for collecting and preparing the data. The article profited from helpful suggestions by Bernadette Witecy and an anonymous reviewer.

Funding

The research described was funded by a grant from the German Research Foundation (DFG) to Martina Penke, Monika Rothweiler, and Markus Hess (UKE Hamburg).

Notes

¹ Hearing aids—provided for mild-to-moderate HI—and cochlear implants—chosen for severe HI—differ fundamentally in their technical procedure and hence in the hearing impression they provide. These differences have been shown to differentially affect language acquisition in HI children supplied with hearing aids or with cochlear implants (Einholz et al. 2015). These two groups of HI children therefore have to be dissociated when investigating language performance.

² There is, of course, no need for a clinical language marker to identify HI, but we need clinical markers for the identification of SLI.

³ The Bonferroni test is a conservative post hoc test that runs pairwise comparisons between the data obtained from different subject groups. It is designed to control the Type I error of assuming a significant effect where there is none, an error that inflates with numerous comparisons. See Field (2013) for further information.

⁴ Welch’s F is a robust technique that corrects the F ratio for unequal variances between groups by adjusting degrees of freedom (see Field 2013).

⁵ The Tamhane test is a conservative post hoc test for data displaying unequal population variances. It runs pairwise comparisons between the data obtained from different subject groups and is designed to control the Type I error that inflates with numerous comparisons. See Field (2013) for further information.

⁶ For TD children, no correlations were calculated as children performed at ceiling level with respect to SV agreement.

Additional information

Funding

The research described was funded by a grant from the German Research Foundation (DFG) to Martina Penke, Monika Rothweiler, and Markus Hess (UKE Hamburg).