Poster #78 - YouTube Kids: Understanding Gender and Emotion through Modern Media

• In Event: Poster Session 02

Abstract

Through emotion socialization, children learn what emotions are, how to express them, and how to respond to them (Eisenberg et al., 1998; Gottman et al., 1996). What children learn about emotions is frequently conflated with implications about gender: stereotypes of masculinity and femininity include emotional display rules (Oliver & Green, 2001). With proliferated online media, the world in which children experience emotions has undergone unparalleled expansion. We endeavored to advance understanding of how videos on the platform, YouTube Kids, contribute to emotion socialization during middle childhood. Previous investigations of television and movies revealed distinct emotion display patterns, with masculine characters displaying more negative externalizing emotions, primarily anger, while feminine characters showed more positive emotions and negative internalizing emotions, including fear and sadness (Oliver & Green, 2001). In the current study, we coded gender and emotion in YouTube Kids videos to determine whether gendered patterns of emotion were present.

We created two ghost users (one in the 5 to 8 years-old group and one in the 9 to 12 years-old group) to select the most popular videos for coding, as indicated by their rankings on the browsing page. YouTube Kids delineates videos into four categories: Reading, Shows, Music, and Explore. Each category suggests 36 videos. We are analyzing the top five videos in each category. Videos were chosen equally from each user, with the fifth video per category being an overlapping video, for a total of 20 videos. User profile creation excludes gender, and the video suggestion algorithm is private. Separate teams are coding gender and emotion. Each video receives an overall code based on all content, including video title and description, as either completely feminine; mainly feminine with some masculinity; equally feminine and masculine; no gender-typed content; mainly masculine with some femininity; or completely masculine. Furthermore, gender presentation of each character is coded as only feminine; both masculine and feminine; neither feminine nor masculine; or only masculine. Emotion coding for each character uses a three-point Likert scale (0-2) to indicate the extent of prototypical emotions such as excitement, happiness, shame/guilt, anger, and fear. To date, 20 videos have been coded for gender, with satisfactory reliability (K = .85 for overall video, K = .84 for characters). Preliminary descriptive findings provide some evidence of egalitarian socialization in YouTube Kids media, as almost half of the videos (9 of 20) were coded as having equally feminine and masculine content (see Figure 1). However, most characters (73%) were depicted with masculine gender presentations. Thus far, five (of the 20) videos have also been coded for emotion (ICCs = .78 - .90 for anger, fear, guilt/shame, happiness, and excitement): characters with varying gender presentations displayed positive emotions, anger, and fear (see Figure 2). Final analyses will include ANOVAs to determine whether creators are disrupting or conforming to gendered emotion stereotypes. This study provides initial data concerning gendered patterns of emotions in prominent kids-focused media: according to preliminary findings, this modern media may function to create a more equal world through egalitarian messages conveyed about emotions and gender.

Authors

• Lauren Taylor Lyles, University of Houston
  Presenting Author

• Leslie Frankel, University of Houston

• Julie C Dunsmore, University of Houston