The Effects of Character-Similarity on Identification and Learning from a Narrative

• In Event: Excellence in Education Research: Early-Career Scholars and Their Work

Abstract

Media pervades the lives of young children: On average, five- to eight-year-olds read or are read to for over 30 minutes and watch almost 1.5 hours of television and DVDs every day (Common Sense Media, 2013). In part, media is pure entertainment: Children read or watch for fun. However, media is also used in school settings, and it has profound educational possibilities both inside and outside school. One factor that may affect children’s learning from media is the level of similarity between the child and a story’s protagonist. Cohen (2001) suggested that similarity of the character to the audience member, in terms such as gender or race, should lead to more identification. Although some research supports this relationship, it has rarely been tested directly. Furthermore, some researchers have suggested that individuals learn more from a narrative when they identify with the main character (Calvert, Strouse & Murray, 2006; Maccoby & Wilson, 1957). These two proposed links, between character similarity and identification, and between identification and learning, suggest that if children’s books or television shows contain characters who are dissimilar to them, they may be less likely to learn from these narratives.

In the current study, 73 primarily White children aged 6 to 8 years were read a
storybook about a child who goes on a field trip to a TV news station, with either a Black protagonist or a White protagonist of their same gender. To test children’s identification with the character, children used visual analogue scales: 10 centimeter lines with labeled endpoints. First, children had training using a line with the words “Black” and “White” at the endpoints and were asked to indicate the color of five squares on the continuum from white to gray to black, in order to help them conceptualize the continuous and relative nature of the scale. On test measures, children indicated their responses on lines labeled with the endpoints “Not at all” and “Very.” Children’s responses were coded for how many centimeters away from the left endpoint they marked. An experimenter read children questions about their engagement in the story and identification with the character, such as “Sometimes when you hear a story, you see yourself as the person in the story. In a funny sort of way, you become the other person. How much did you see yourself as the person in the story?” (adapted from Jose & Brewer, 1984). To test
memory and learning, children were first given an open-ended free recall period in which they were asked to describe everything they remembered from the story. Then, they responded to targeted questions about the story content. Children also completed the PPVT as a standardized measure of verbal ability.

Individual regression models were conducted predicting identification and comprehension measures from character race. All models tested for effects of, and if needed, controlled for other variables of interest, specifically age, gender, and verbal ability. Estimates of effects were obtained from a full model (including all factors, interactions, and covariates that might be of interest) and compared to simpler, nested models. The results showed that children were more likely to strongly identify with the

White character (M = 5.3, SD = 3.6) than with the Black character (M = 3.7, SD = 3.7), controlling for gender, B = 1.6, p = .053, d = .44. Furthermore, although children did not answer more comprehension questions correctly after seeing the story with a White character, p =. 57, they did freely recalled more details from a story with a White character (M = 27.3, SD = 17.1) than a Black character (M = 19.7, SD = 12.5), controlling for age, gender and PPVT score, B = 8.4, p = .005, d = .51.

One explanation for this finding is that children who saw a story with a White character may have thought of the story as more relevant to their own lives, and thus encoded and recalled more details without specific prompts. Another possibility is that children were more motivated to report all the details they remembered in the same race story. These findings – especially if replicated with children of color – have the potential to inform our understanding of children’s comprehension of narrative texts and the effect of character race on comprehension. More broadly, this research has important implications for children’s educational media.

Author

• Rebecca Dore, University of Delaware