Minute-by-minute variations in parental number talk and their implications for coding parent-child interactions

• In Event: Poster Session 12
  In Poster Session: PS 12 Topic - Parenting & Parent-Child Relationships

Abstract

Parental language input during dyadic interactions, whether domain-general (e.g., total number of utterances) or domain-specific (e.g., number talk), are consistent predictors of children’s academic skills (e.g., Elliott et al., 2017; Reynolds et al., 2019). For instance, Elliott et al. (2017) found that parents’ talk about numbers greater than 10 was positively associated with 5- and 6-year-old children’s math ability. However, across previous studies, the length and context of observed parent-child interactions varies widely. These differences may be driven by an array of methodological considerations, such as the nature of the activities recorded (e.g., goal-oriented, free play, natural) and how old the target children are. Given the constraints on time and resources to reduce data for later analyses, the current study investigated how the frequency of parental talk changes within tasks and the utility of coding additional minutes of talk. Specifically, we examined variations in the average amount of number talk within 1-minute chunks of time during three different structured parent-child interactions. Our goal was to understand a potential way to reduce the cost and burden of data collection and processing on families and research teams.

Parent-child dyads (n=122, mean child age=4.41 years, SD=.31) were video-recorded interacting with a wordless picture book, grocery store toys, and a magnetic puzzle board for about 5 to 8 minutes each. Videos were transcribed verbatim and parental number talk was coded at the utterance level. Given that dyads varied in the amount of time that they engaged in each task, we report analyses with different sample sizes for each activity, including all dyads who interacted for at least 6 minutes with the book (n=73), grocery store toys (n=124), and puzzle board (n=103). We found that the mean frequency of number utterances generally increased with each additional minute in the book and grocery tasks, while remaining relatively constant in the puzzle task (Table 1). Further, we conducted six-step hierarchical linear regression analyses relating total number talk to segmented minutes of talk in each task. We began by adding the first minute of number talk in the first models and then included an additional minute of talk with each step, ending with the sixth minute segment. Each additional minute of number talk explained significantly more variance than the previous steps, with the exception of the first two minutes of the book task. The ΔR2 for each minute ranged from .14 - .24 (book), .09 - .23 (grocery), and .05 - .14 (puzzle).

Our descriptive findings suggest that the flow of parental number talk across time is partially task-dependent. It is possible that average number talk increased with time in the book and grocery activities, because dyads first had to familiarize themselves with the tasks but once this was achieved, these activities were more conducive to talk about number. However, our regression results suggest that it is worth transcribing and coding the entirety of parent-child interactions as every added minute explained a substantial amount of additional variance in parental number talk.

Authors

• Shirley Duong, University of Pittsburgh
  Presenting Author

• Leanne Elliott, University of Pittsburgh
  Non-Presenting Author

• Heather J. Bachman, University of Pittsburgh
  Non-Presenting Author

• Elizabeth Votruba-Drzal, University of Pittsburgh
  Non-Presenting Author

• Melissa E Libertus, University of Pittsburgh
  Non-Presenting Author