Poster #189 - Gendered Robots Can Change Children’s Gender Stereotypes

• In Event: 2-163 - Poster Session 08
  In Poster Session: PS 08 Section - Sex, Gender

Integrative Statement

Research in social robotics and human-robot interaction has identified gender as an important factor of robot design. Female and male gendered robots can illicit different levels of likeability, acceptance, information seeking, and cooperation from adults, especially when fulfilling gender stereotyped roles (Kuchenbrandt et al., 2014; Tay et al., 2014). For these reasons, some researchers suggest designers should build gender cues into social robots to improve rapport, engagement, and marketability (Sandygulova & O’Hare, 2018). Indeed, many popular robots already in use are female gendered and fulfill stereotypical female roles (e.g., Apple’s Siri and Google’s Alexa; Devlin, 2018). However, little is known about the effects of gendered robots on children’s thinking. Children treat robots as informants and form meaningful social and moral relationships with robots (Breazeal et al., 2016; Kahn et al., 2012). Thus, children may treat gendered robots as models of gender norms in a manner similar to characters in media and gaming environments (e.g., Coyle & Liben, 2016). The present research examined whether female gendered robots fulfilling stereotypical and counter-stereotypical jobs affect children’s occupational gender stereotypes.

Forty-three 6- to- 10-year-old children viewed short animated and narrated cartoons about three female gendered robots currently in research and development. Each robot had a unique occupation (e.g., scientist or nurse), interest (e.g., drawing cars or baking cookies), and personality (adventurous/smart or loving/gentle). Children in a Counter-Stereotype condition viewed masculine robots whereas children in a Stereotype-Reinforcing condition viewed feminine robots (see Figure 1). Culturally stereotyped masculine and feminine qualities were adopted from the Children’s Occupation, Activity, and Trait-Attitude Measure (COAT-AM; Liben et al., 2002). Before and after viewing the videos, children completed a 12-item version of the Occupation Gender Stereotype measure of the COAT-AM (e.g., “Who should be a florist? Only men, only women, or both men and women?”). Children additionally completed an analogous measure that assessed occupational stereotypes for robot targets (e.g., “Who should be a plumber? Only Jack-Bot, only Jill-Bot, or both Jack-Bot and Jill-Bot?”).

Did exposure to gendered robots change children’s gender stereotypes? As can be seen in Figure 2, exposure to Stereotype-Reinforcing robots increased children’s occupational stereotyping for human targets (e.g., saying only women could be florists), OR = 1.71, 95% CI [1.06, 2.74], whereas exposure to Counter-Stereotype robots had little effect, OR = .80, 95% CI [.50, 1.28]. In contrast, Counter-Stereotype robots decreased children’s stereotyping of robot targets (e.g., saying only Jack-bot could be a plumber), OR = .51, 95% CI [.32, .81], whereas Stereotype-Reinforcing robots had little effect, OR = .96, 95% CI [.60, 1.53].

These results demonstrate robots exhibiting cultural gender stereotypes can affect children’s gender stereotypes. Critically, our findings suggest gendered robots might serve to both reinforce and undermine potentially harmful gender stereotypes in children. This work suggests designers and consumers need to carefully consider the implications of gendering robots for children. Moreover, this research highlights the theoretical and applied contributions of studying child-robot interaction as a domain of child development.

Authors

• Kallyn Song-Nichols, Occidental College
  Presenting Author

• Andrew G Young, Occidental College
  Non-Presenting Author