This collaborative project brought together a team of three research sites to examine the contextual (parents’ and teachers’ perceptions of the role of technology in science education; available options in the media landscape) and social-cognitive (children’s perceptions of video characters and intelligent agents) factors that support learning from educational video and interactive media in early childhood. Collectively, several main findings emerged from this project.

First, about 2/3 of parents recognize that both science and mathematics are relevant and advantageous to their child in the future. Parents also believe that children can learn STEM concepts from digital media. More than 2/3 of respondents believe that their child learns about science from television, DVDs, and apps on mobile devices, while approximately 3/4 of respondents believe that their child learns about math from television, DVDs, and online video tutorials. That being said, there are differences in the representation of science and math topics on preschool children’s programming, with some topics like working with numbers being overrepresented compared to other topics, such as natural hazards. In addition, parents of STEM career families report their children use less science and math media compared to parents of non-STEM career families. Additionally, parents of non-STEM career families who use the most media report their children have significantly worse science and math skills.

Second, between the ages of 18 months and 2 years, children who have formed an emotional bond with an on-screen character (especially if they have a toy that represents that character) will be more likely to learn foundational math skills from that character. As children become 3- to 6-years-old, their learning from video characters continues to be driven by social-cognitive factors, such as trusting that an animated child character is a relevant source of information for how to solve a difficult problem. Eye-tracking studies found that 4- to 5-year-old children looked more at characters who exerted high effort to solve problems, and also imitated more from those characters; in contrast, 5- to 6-year-olds looked more at characters they perceived to be expert problem solvers, and also transferred more from those characters. Gender also plays an important role during these years: we have found that preschoolers do have gender-based stereotypes for familiar television characters, with male characters seen as being more interested in science than female characters. In addition, boys and girls are more likely to identify with characters that corresponded with their perceived gender. In sum, when presented with educational video, children are relying on their prior experiences with the characters they see as well as their existing gender stereotypes to determine whether the character is sharing relevant information, whether to attend to that information, and whether to use that information in a novel setting.

Third, learning from interactive media demonstrates can also build on social-cognitive mechanisms of learning, however, our newest findings suggest the need for much more research into the mechanisms of learning through interactive games in early childhood. Our studies have found that having an interactive intelligent agent who responds in meaningful and contingent ways with a child learner promotes early math learning. However, interactivity through tablet play contributed to better performance on near transfer tasks, while participants in a non-interactive condition performed better on far transfer tasks.

Last Modified: 11/26/2018
Modified by: Rebekah Richert