Effects of stimulus age on the endorsement of gender stereotypes

• In Event: Intersectional Perspectives on the Development of Social Cognition

Abstract

Gender and age are salient social categories from early in development. By 2- to 3 years of age children’s gender stereotypes reliably govern their expectations of others behavior (e.g., Ruble et al., 2006), and children express beliefs that some aspects of appearance, beliefs, and behaviors vary between children and adults (e.g., VanderBorght & Jaswal, 2009). However, whether children’s beliefs about gender and age intersect, such that gender stereotypes might be expressed differently when asked about children, as compared to adults, has not been investigated. Filling this gap in knowledge has conceptual and practical consequences by shaping our understanding of children’s ability to integrate information across social domains in an intersectional manner, and by informing future research and helping researchers select stimuli that are most appropriate—and most likely to pick up on—the construct or beliefs under investigation.
Here, in a pre-registered study (N = 401), we examined if young children (3.0 – 6.9-year-olds, Mage = 5.03 years, n = 145) and adults (n = 256) were more likely to express gender stereotypes when presented with child or adult stimuli. Participants heard 20 questions about behavioral and psychological properties, each tapping a different gender stereotype (e.g., “Which of these people really likes to play with dolls?”). Participants selected their response (male or female) for each question by selecting between four child faces (2 White boys, 2 White girls). Participants completed a second block with identical questions but where 4 adult faces (2 White men, 2 White women) were presented; block presentation order was counterbalanced. We summed the number of stereotype-consistent responses across each block separately and used generalized linear mixed effects models to examine whether the probability of giving a stereotypical response varied as a function of target age (child; adult) or participant age group (children; adults), while adjusting for presentation order. We found no effect of participant gender on responses, so it was excluded from analyses.
Overall, both children (M = 0.59, 95% CI(0.57, 0.60), t = 12.22, p < .001) and adults (M = 0.60, 95% CI(0.59, 0.61), t = 15.69, p < .001, Fig. 1) expressed gender stereotypes, and, in children, expression of stereotypes increased with age, ß = 0.85, SE = 0.21, z = 3.97, p < .001. However, we found no effect of target age for either children or adults, suggesting that participants’ gender stereotypes did not generally intersect with age. This pattern was largely similar when we examined properties individually, although we found some suggestion that adults (but not children) were sensitive to both age and gender content for child-typed beliefs (e.g., playing with dolls: interaction of target age and participant age group, ß = -1.71, SE = 0.44, z = -3.91, p < .001; Fig. 2). These results suggest that young children’s gender stereotypes emerge early and strengthen across development, and that children apply gender stereotypes equally to child and adult stimuli. Future research should explore whether children show this undifferentiated pattern when both gender and age are made salient and directly contrasted.

Authors

• Tara Mandalaywala, University of Massachusetts Amherst
  Presenting Author

• Alix Alto, City University of New York Graduate Center
  Non-Presenting Author