Show or Tell: Children’s Learning About Food From Action vs. Speech

• In Event: Children’s Utilization of Shared Information Across Contexts

Abstract

Many interventions to promote healthy eating and prevent obesity deliver verbal lessons to children in preschool classrooms but have observed modest beneficial effects (Colquitt et al., 2016; Gripshover & Markman, 2013; Sigman-Grant et al., 2014; Wolfenden et al., 2012), highlighting an opportunity to better understand whether different modalities of information sharing may influence children’s food choices. While many studies highlight that preschool-aged children are sophisticated social learners (Harris & Koenig, 2006; Nguyen et al., 2012; VanderBorght & Jaswal, 2009), it is unknown how different types of information (e.g., action vs. verbal testimony) influence children’s social learning. Therefore, the present study examines the relative effectiveness of showing vs. telling children about food.

Three- to 6-year-old children (n=71) watched videos of two people eating apple-broccoli puree (selected as an unfamiliar food). One person ate 5 bites of food; the other said they liked the food. Children were offered the foods from the videos, had up to 5 minutes to eat the foods, and then rated the foods from really yucky to really yummy. Children also completed an analogous toy task and novel object task. Children were randomly assigned to task order (food, toy, or object first). Parents completed the Child Eating Behavior Questionnaire (Wardle et al., 2001), a validated questionnaire about children’s eating habits and pickiness.

Children did not differentiate between the two types of information provided: They ate similar amounts of the food they saw someone eat (M = 4.86 g) and the food they heard someone talk about (M = 4.47 g), F(1, 67) = 0.13, p = .719. No effect of task order or interaction were observed (ps > .5). A similar pattern was observed for children’s taste ratings: No significant effect of information type (act: M = 1.66; talk: M = 1.80), task order, or interaction were observed (ps > .5). Parents’ responses on the Food Fussiness subscale of the CEBQ were associated with children’s food intake (b=-4.53, SE=1.08, t=-4.21, p<.001) and ratings (b=-0.40, SE=0.17, t=-2.37, p=.020); the pickier the child, the less they ate and the worse the rated the foods. Children also did not differentiate between the two toys based on the way information was provided (p=.219) and rated both toys as fun (ps<.001). When asked which actor used the unfamiliar object the right way, 48 selected the action testimony and 19 selected the verbal testimony (binomial test: p<.001).

This study failed to observe differences in children’s response to action and verbal testimony when children were presented with unfamiliar foods. We did find an action advantage in the novel object task, but this was just one measure out of many in the study. These association with the CEBQ provides converging evidence that parent ratings of children’s eating behavior at home align with observations of children’s eating behavior in laboratory settings. Taken together, these findings raise important questions as to what constitutes a learning context and the implication of small effects or resistance to try unfamiliar foods for the study of children’s social learning about foods.

Authors

• Jasmine M. DeJesus, University of North Carolina at Greensboro
  Presenting Author

• Shruthi Venkatesh, University of North Carolina at Greensboro
  Non-Presenting Author