Assessing the Language Abilities of Preterm-Born Children: An Examination of Standardized Testing and Language Sample Analysis

Abstract

Purpose: To understand how best to assess the language abilities of preterm-born children, this study: (a) compared preterm- and term-born children's language skills using standardized testing and language sample analysis (LSA), (b) investigated how executive function skills and the language sampling context respectively affect standardized test and LSA scores, and (c) examined the pattern of associations between standardized test and LSA scores among preterm-/term-born groups. Method: Twenty-five term-born and 23 preterm-born 2-year-old singletons were administered the language scales of the Bayley Scales of Infant and Toddler Development–Third Edition (receptive communication, expressive communication, language composite scores). Parent–child free-play recordings were used to quantify the (para)linguistic features of the children's speech. Executive function was measured via parent report. Results: The preterm-born group obtained significantly lower scores than the term-born group on all Bayley language measures (though differences were not consistently observed when using cutoff scores). Few preterm–term differences in LSA measures were found. The preterm–term differences in Bayley scores were not explained by between-group differences in executive function. Some preterm–term differences in LSA scores were moderated by the language sampling context. The preterm- and term-born groups exhibited different patterns of Bayley–LSA correlations. Conclusions: Preterm language difficulties were more apparent on standardized test than LSA scores. Nonetheless, the Bayley–LSA correlations indicate that poor test performance (linked with preterm birth) is associated with functional communication difficulties. The discussion outlines the complementary utility of standardized tests and LSA while acknowledging the limited utility of cutoff scores and the confounding influence of the language sampling context. Supplemental Material: https://doi.org/10.23641/asha.26142661