Lilliput: speech perception in speech-weighted noise and in quiet in young children

Abstract

Objective

The aim of this study was to develop an open-set word recognition task in speech-weighted noise and in quiet for young children and examine age effects for open versus closed response formats.

Design

Dutch monosyllabic words were presented in quiet and in stationary speech-weighted noise to 4- and 5-year-old children as well as to young adults in an open-set response format. Additionally, performance in open and closed context was assessed, as well as in a picture-pointing paradigm

Study sample:

More than 200 children and 50 adults with normal hearing participated in the various validation phases.

Results

Average fitted speech reception thresholds (50%) yielded an age effect between 4-year and 5-year olds (and adults), both in speech-weighted noise and in quiet. The closed-set format yielded lower (better) SNRs than the open-set format, and children benefitted to the same extent as adults from phonetically similar words in speech-weighted noise. Additionally, the 4 AFC picture-pointing paradigm can be used to assess word recognition in quiet from 3 years of age.

Conclusions

The same materials reveal performance differences between 4 and 5 years of age (and adults), both in quiet and speech-weighted noise using an open-set response format. This relatively small yet significant difference in SRT for a gap of only 1 year shows a developmental change for word recognition in speech-weighted noise and in quiet in the first decade of life.

The study is part of the protocol registered on ClinicalTrials.gov (ID = NCT04063748).

Keywords:

• Speech perception in quiet
• speech perception in noise
• open-set
• closed-set
• children

Acknowledgements

The authors gratefully thank all the children who participated in our research project and, thereby, contributed to the development of the Lilliput speech materials. Moreover, the many different phases of this study would not have been possible without the dedication of our Master students of Speech Pathology and Audiology Sciences KU Leuven, who tested the children in the framework of their Master thesis.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by a TBM-FWO grant from the Research Foundation-Flanders [grant number T002216N] and H2020-MSCA-ITN-2019 Comm4Child [grant number 860755].