Measurement of Lexical Diversity in Children’s Spoken Language: Computational and Conceptual Considerations

Abstract

BackgroundType-Token Ratio (TTR), given its relatively simple hand computation, is one of the few LSA measures calculated by clinicians in everyday practice. However, it has significant well-documented shortcomings; these include instability as a function of sample size, and absence of clear developmental profiles over early childhood. A variety of alternative measures of lexical diversity have been proposed; some, such as Number of Different Words/100 (NDW) can also be computed by hand. However, others, such as Vocabulary Diversity (VocD) and the Moving Average Type Token Ratio (MATTR) rely on complex resampling algorithms that cannot be conducted by hand. To date, no large-scale study of all four measures has evaluated how well any capture typical developmental trends over early childhood, or whether any reliably distinguish typical from atypical profiles of expressive child language ability.Materials and MethodsWe conducted linear and non-linear regression analyses for TTR, NDW, VocD, and MATTR scores for samples taken from 946 corpora from typically developing preschool children (ages 2–6 years), engaged in adult-child toy play, from the Child Language Data Exchange System (CHILDES). These were contrasted with 504 samples from children known to have delayed expressive language skills (total n = 1,454 samples). We also conducted a separate sub-analysis which examined possible contextual effects of sampling environment on lexical diversity.ResultsOnly VocD showed significantly different mean scores between the typically -developing children and delayed developing children group. Using TTR would actually misdiagnose typical children and miss children with known language impairment. However, computation of VocD as a function of toy interactions was significant and emerges as a further caution in use of lexical diversity as a valid proxy index of children’s expressive vocabulary skill.DiscussionThis large scale statistical comparison of computer-implemented algorithms for expressive lexical profiles in young children with traditional, hand-calculated measures showed that only VocD met criteria for evidence-based use in LSA. However, VocD was impacted by sample elicitation context, suggesting that non-linguistic factors, such as engagement with elicitation props, contaminate estimates of spoken lexical skill in young children. Implications and suggested directions are discussed.

Personal Narratives of School-Age Children: A Cross-Sectional Developmental Study

“…The obtained results are consistent with previous studies investigating the personal narratives in younger children (aged 5-8 years) from English-speaking countries, which also showed a gradual increase in syntactic complexity and, in contrast to our results, also in lexical diversity [26,37]. These contrasting findings may be explained by different age groups involved in the studies and by the different measures used to evaluate lexical diversity, with one of the measures used in the current study (LTR) known to be less sensitive to age [53] and affected by sample length [43]. Regarding the latter, it must be mentioned that the narratives were on average rather short, and in some children, this was more evident than in other children.…”

“…TTR and LTR also tend to decline as the TNW increases because of more frequent use of common words, such as frequently used grammatical function words. Despite the shortcomings, both NDW and TTR are reliable under conditions typically used with older children, such as elicited narrative or story retell, in which prompts remain stable from sample to sample [43]. In this study, NDW and LTR were chosen as measures of lexical diversity.…”

“…It remains unclear whether this is something related particularly to this type of narrative or whether other elicitation prompts and methods may encourage children to produce longer personal narratives. Some of the measures used in the study (especially LTR) are very sensitive to sample length (see [43]). Therefore, the choice of this measure over others and its calculation on samples of different lengths are also weaknesses of our study.…”

Measuring Language Development From Child-centered Recordings

“…These measures are vocabulary diversity (VOCD), mean length of utterance (MLU) in morphemes, and the Index of Productive Syntax (IPSyn). VOCD is reputedly the best measure of lexical diversity for several reasons, including its relative independence from transcript length and the fact that lexical diversity is done on lemmas rather than surface forms [18]. MLU measures the average length of sentences, and is thought to reflect both morphological and syntactic development.…”