Predictive Strength of Auditory Maturity Across Different Levels of Language Ability: An Exploratory Quantile Regression Study.

Abstract

Auditory evoked potential-age (AEP-age) is proposed to index auditory maturation and has been found to predict language skills in children with and without a language disorder. However, reporting average effects using linear regression does not fully capitalize on the potential of AEP-age to estimate individual differences in young children. This study used a quantile regression approach to examine the predictive utility of AEP-age for 105 typical and neurodiverse 7-10-year-old children (61 males; 44 females; largely monolingual English) with varying language skills without creating subgroups. Although linear regression did not find an association between AEP-age and language skills, the quantile model added specificity by revealing differential associations. AEP-age was only related to language skills for children at around the median point of the language continuum, but, not for those at the lowest or highest end of the language distribution. Overall, the quantile regression methodology provides us with the flexibility of understanding how AEP-age is related to different language abilities.