Early literacy in the digital age: Relationships between visual attention, digital exposure and emergent literacy

The emergence of digital technology has brought about children being exposed to digital text from a very young age. To explore how this potentially shapes reading development, there is the need to investigate whether it is necessary to identify digitally-relevant predictor skills of reading alongside other more traditional literacy precursor skills. While phonological decoding has repeatedly been shown as a strong predictor of reading success (Snowling, 2000; Vellutino et al., 1991), visual attention has only more recently been recognised as important moderator of reading performance (Franceschini et al., 2012). With an increasingly rapid move away from standardised typographic text formats (Picton, 2014), individual variation in visual attention skills is likely to be an important moderator of reading performance. This study explored longitudinal relationships between children’s exposure to digital devices, visual attention development, traditional predictors of reading and their relative impact on word reading after the first school year. UK children from a normative group of emerging readers (N = 140) were recruited at the beginning of Reception (4;7 years) and reassessed at the start of Year 1. Participants were tested on their visual attention and other traditional literacy precursor skills, including: phonological awareness, letter-sound knowledge, receptive vocabulary, short-term memory, and non-verbal ability. Further, information was gathered on children’s digital exposure. Cross-sectional data at both time points demonstrated moderate correlations between visual attention and traditional precursor skills of reading, but no relationships between children’s digital exposure and their visual attention. Within regression analyses, visual attention, while contributing variance longitudinally to single word reading after one year of schooling, was not found to be a unique predictor of single word reading after other variables were entered into the model in a prior step. The relevance of these findings, as well as related theoretical and practical implications are discussed.