Optimizing language environment measures: combining LENA audio recordings with the Q-BEx questionnaire

The amount of language input that children receive is an important predictor for their language development (Anderson et al., 2021; Hart & Risley, 1995; Huttenlocher et al., 1991; Lieven et al., 2019; Rowe et al., 2012). For bilingual children, the quantitative language input can differ strongly per language. The quantitative language input depends on different factors, such as the languages that each parent speaks, how talkative parents are, whether the child is going to school, whether the child receives input from siblings, etc. These factors vary per bilingual family (Hoff & Core, 2013), making it difficult for researchers to capture the exact language input per language for bilingual children. The present study compares currently used methods that quantify bilingual language input and explores a new method, with the aim to optimize quantitative language input measures in children’s bilingualism research. We focus on the use of parental questionnaires, specifically the Questionnaire for Quantifying Bilingual Experience (Q-BEx; De Cat et al., 2022), and daylong audio recordings made by the Language Environment Analysis (LENA). The new method is a combination of the questionnaire and audio recording data: it combines the observed language use from the daylong audio recording with the general weekly contact hours gathered with the Q-BEx questionnaire. This new, combined, method has been developed assuming that it measures quantitative language input more accurately than questionnaires and audio recordings individually.

We investigate the three different language environment measures by correlating them with children's vocabulary scores. Our sample consists out of 52 bilingual children (Turkish/Dutch or Polish/Dutch) aged 3-5 years old.