Using latent variable analysis to capture individual differences in bilingual language experience

Abstract

Abstract Bilingualism is an experience that varies across a continuum and can change across the lifespan. Psychometric research is an underexplored avenue with the potential to further our understanding of the mechanisms and traits underlying bilingual experiences. Here, we developed and validated a social network questionnaire to measure sociolinguistic features in 212 individuals via personal social network. Confirmatory factor analysis examined the measurement structure of the variables. Compared to a one-factor model, the best fitting model was a two-factor model in which the language experience of the individual (i.e., ego) and the language experience of the individual's network (i.e., alters) were correlated latent factors under which aspects of the bilingual experience loaded. Additional analyses revealed other potential ways to examine the data in future analyses. These results provide the first measurement model of bilingual experiences, and provide support for theoretical accounts suggesting differential neuropsychological outcomes based on individual bilingual variability. The results also support the use of social network tools to capture differences in bilingualism.