Moment-to-Moment Digital Media Use Associated with Likelihood and Quantity of Parent-Child Language Interaction

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Prevention and Interventions

Abstract

Media use by young children is common and likely has increased during the global Covid-19 pandemic (Vanderloo, et al., 2020). While early media use is negatively associated with attention (Courage & Setliff, 2010) and other executive functions (Linebarger, et al., 2014), it is also associated with decreases in rich parent-child interaction (Lavigne, et al., 2014). Early opportunities for language stimulation, such as the number of adult words children hear and the number of back and forth conversations with caregivers, are robustly related to children’s later language ability and academic achievement (Gilkerson, et al., 2018; Greenwood, et al., 2019). These opportunities are less frequent when there is audible television or digital media present in analyses of daylong language recordings and in observational studies (Christakis, et al., 2010; Pempek, et al., 2014). However, using daylong recording estimates may miss nuances in smaller time intervals and brief observational studies may overlook larger patterns. The current study uses daylong automatic recordings to investigate the relationship between electronic media and child-directed speech and conversational turns in 5-minute segments.

Participants were 105 primary caregivers from families with children between 1 and 30 months of age (M = 11.34 months) who were part of an intervention study to increase parental provision of language to children. Baseline data was used in these analyses, which was collected in 2017-18. Daylong audio recordings and estimates of adult word count and conversational turns were provided by the Language Environment Analysis System, which is shown to be reliable and valid (Greenwood, et al. 2018). Data were exported into 5-minute segments and evaluated for the relationship between adult word count, conversational turns, and media heard per 5 minutes using zero-inflated generalized linear mixed models (GLMM), controlling for child age and parent educational level.

Results from the GLMM of media predicting adult word count showed the amount of electronic media per 5-minutes was significantly negatively associated with the number of adult words to a child in any 5-minute segment (risk ratio = -0.004, Wald z = -8.91, p < .001). Adult words in a 5-minute segment with no media was predicted at 80.48 (SE = 3.5) and with 300 seconds of media, was only 26.76 (SE = 3.0) words. Critically, 5-minute segments with any media were 56% more likely to include no adult talk. In the GLMM for conversational turns per 5 minutes, the same relationship was true (risk ratio = -0.007, Wald z = -13.13, p < .001). Conversational turns per 5-minute segment with 0 media was predicted at 2.15 (SE = .11) and with 300 seconds of media, predicted at only .25 (SE = .04). Five-minute segments with any media were 32% more likely to have no conversational turns.

While early childhood professionals have long been cautioning adults to restrict children’s media use, the practice continues. Shifting the focus to increasing the amount of adult talk – even when using media, may be a more sustainable approach. Future directions include using data and targeted interventions to teach caregivers to mediate children’s media use.

Authors

• Craig K. Van Pay, University of Oklahoma
  Presenting Author

• Constance C Beecher, Iowa State University
  Non-Presenting Author