An Early Lexical Screening Tool for British English: Psychometric Properties and Clinical Utility.

Abstract

Background: Existing direct measures of vocabulary for ages 2-3 years are insufficient as vocabulary increases to include nouns, verbs, adjectives and adverbs, reflecting qualitative changes that support emerging grammatical abilities.

Aims: The present study aims to establish the psychometric properties and validate a lexical screening tool for 19-36-month-old British English-speaking children, the WinG (Words in Game) test. Specifically, it examines gender differences in internal consistency, concurrent validity and predictive gender invariance across comprehension and production tasks. In addition, the study evaluates the diagnostic accuracy of the tool by establishing clinical cut-off scores to distinguish children with typical versus at-risk language development. Finally, it presents developmental trends in toddlers' aggregate lexical comprehension and production scores, reported in percentiles and stratified by gender.

Methods and procedures: Participants were 336 English-speaking children aged 19-36 months. Sub-set of 85 children were screened with the Words in Game (WinG) test and the Preschool Language Scale (PLS-4). The PLS-4 scores were used as reference tests to ascertain psychometric properties and estimate the sensitivity and specificity of the WinG comprehension and production tasks.

Outcomes and results: Girls outperformed boys on both comprehension and production tasks. The WinG comprehension and production tasks showed strong concurrent validity, with scores positively and highly correlated with the PLS-4 reference measures. Predictive gender invariance was also demonstrated, as both WinG dimensions predicted PLS-4 scores equivalently across genders. Using the 15th percentile as a cut-off, the WinG test demonstrated fair sensitivity for comprehension and good sensitivity for production, though specificity was below the desirable threshold for both tasks.

Conclusions and implications: The tool provides a valuable starting point for identifying early receptive and expressive lexical difficulties through direct assessment by speech and language therapists (SLTs). Clinical implications and limitations are also discussed.

What this paper adds: What is already known on this subject Lexical receptive and expressive abilities emerge and expand rapidly in the second and third years of life. Clinicians and Speech and Language therapists (SLTs) need a systematic, reliable tool to observe the early lexical skills of toddlers. What this study adds to existing knowledge This study established the psychometric properties and validated the Words in Game (WinG) test from a large sample of 19-36-month-old British English learning children to measure the receptive and expressive nouns and predicates (verbs, adjectives and adverbs). The PLS-4 scores were used as a reference standard test. The WinG comprehension and production tasks showed strong concurrent validity, with scores positively and highly correlated with the PLS-4 reference measures. The developmental trends in toddlers' lexical comprehension and production scores are reported in percentiles and stratified by gender. The diagnostic accuracy of the WinG test was ascertained, performing receiver operating curves (ROC) on comprehension and production WinG tasks, entering the 15th percentile clinical cut-off scores, as indicated by the PLS-4 reference measures. What are the clinical implications of this study? The WinG test is a valid tool designed to screen the comprehension and production of words of British English children in the second and third years of life. The WinG test is a tool to be administered by professionals with expertise in language development, such as SLTs. Thanks to its systematic, direct observation approach and validation against established measures, the WinG test demonstrates fair psychometric properties. It provides SLTs with a reliable tool for early screening of toddlers' lexical skills to identify language delays or disorders.