Investigating language exposure and screen time in the early home environment: A longitudinal analysis of a prospective cohort study

Abstract

The acquisition of language is one of the most critical developmental milestones for children during their early years. Despite this, a significant number of children are starting school without the speech, language, or communication (SLC) skills required to facilitate their learning and development, setting them on a trajectory of poorer academic, mental health, and employment outcomes into adulthood. This has led to calls from experts in the field to consider poor early language development a public health problem. Shifting to an early intervention approach would alleviate the current burden on clinical speech and language services and instead provide young children access to universal and targeted services that aim to promote language development. One mechanism that has been proposed as an effective intervention strategy is increasing children’s exposure to language to promote a language-rich home environment. Therefore, the focus of the current thesis is understanding Australian children’s early home environments to inform a public health approach to support early language development. This thesis used data collected in an Australian prospective cohort study which utilised advanced speech recognition technology to objectively measure the home language environment among children aged between 6 and 48 months old. Four research studies were undertaken within this thesis that aimed to (1) characterise a child’s early home language environment, as measured by the amount of parent-child talk, (2) understand whether there are socioeconomic differences in the early home language environment, (3) investigate the causal impact of parent-child interactions throughout the early years on language development at 36 and 48 months, and (4) explore the relationship between children’s screen time and early language exposure to help inform future intervention strategies. The first two studies took a descriptive epidemiology approach to understand the early home language environment as characterised by the number of adult words, child vocalisations, and conversational turns (referred to collectively as parent-child talk) a child is exposed to across their first 18 months of life, and how this differs by levels of maternal education (as a measure of socioeconomic status). Findings from the first study indicated no differences between high and low maternal education groups within the first year of life but considerable variation across families. The second study, however, demonstrated that a significant difference between maternal education groups emerges by 18 months old, with children categorised into the high-educated group hearing more adult words and engaging in more interactions than their peers from low-educated backgrounds. The third study employed the use of causal inference methodology to examine the impact of parent-child interactions throughout the first four years of life on children’s language development. This study demonstrated a small causal effect when increasing the number of parent-child interactions across the early years on children’s language development at 3 and 4 years old, after controlling for time-varying exposures and confounding. The fourth study explored the concept of ‘technoference’ which posits that screen time interferes with opportunities for talk and interaction. Using longitudinal mixed effect models, with data collected when the children were 12 to 36 months old, this study demonstrated that increases in screen time reduced the number of adult words and parent-child interactions a child is exposed to during the first three years of life. Overall, findings from this thesis demonstrate the importance of a language-rich home environment to support language outcomes for young children. To ensure early intervention is effective and equitable, a public health approach is required, given the considerable variation in young children’s home environments and the emergence of socioeconomic differences in language exposure by 18 months old. When considering the implementation of interventions, the findings suggest a proportionate universalism approach may be most appropriate whereby universal services are available to all families but designed with a scale and intensity proportionate to the level and type of disadvantage. Interventions that aim to promote early language may also benefit from strategies that educate parents on how to reduce their family’s screen time to create more opportunities for daily language interactions.