Poster #241 - Parasocial Relationships with a Media Character Improves Children’s Early Math Skills

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Technology, Media & Child Development

Integrative Statement

Introduction: In the 21st century, intelligent agents will be able to interact with children, responding contingently to what children say. Educational media characters, who permeate young children’s lives, hold particular promise as intelligent agents, as many children form close emotional bonds with them, known as parasocial relationships (Calvert, 2015). This study examined the impact of a prototype of the popular character, Dora the Explorer, on children’s learning of the add-one rule, that adding one to a number increases the total by one unit. Faster, accurate response times on add-one problems provide a foundation for more advanced math skills (Baroody, Purpura, Eiland, & Reid, 2015).

Method: Four-year-old children (n = 95; Mage = 4.84) played the Dora game in Study 1 (Feasibility study with 4 rounds; n = 38; 22 girls), and Study 2 (Final Prototype with 3 rounds; n = 57; 28 girls). In both studies, children initially answered the attachment and friendship subscale (friend, trustworthy, safe, cute) of the child parasocial relationship measure on a 5-point Likert scale (Cronbach’s α = .70; Richards & Calvert, 2017). Next children played the game using a Wizard of Oz procedure in which a hidden adult responded contingently to children’s verbal interactions with Dora through an onscreen interface.

Using session videos, research assistants assessed children’s latency scores (average amount of time in seconds to solve each math problem correctly; Cronbach’s αs = .98 - 1.00), math talk interactions (one point for any numerical reply to Dora’s math queries; Cronbach’s αs = .95 - 1.00), and small talk interactions (one point for each appropriate reply to personal questions, e.g., favorite color; Cronbach’s αs = .96 - .99). Parasocial relationship scores were computed as the average of children’s attachment and friendship scores.

Results: OLS regression analyses were computed on average latency scores on add-one problems with average attachment and friendship scores as the main predictor, controlling for math talk and small talk. The overall model was significant in Study 1, F (3, 32) = 6.62, p < .01, adjusted R2 = 0.31, and Study 2, F (3, 53) = 7.12, p < 0.001, adjusted R2 = 0.29. In Study 1, for each additional point higher children reported on attachment and friendship scores, they answered math problems correctly almost 7 seconds faster (p = 0.006; see Figure 1). In Study 2, for each additional point higher children reported on attachment and friendship scores, they answered math problems correctly almost 3.5 seconds faster (p = .002; see Figure 2).

Conclusion: As children’s meaningful parasocial relationships with the character increased, children answered add-one problems more quickly and accurately. These findings have human-computer interaction design implications, with socially meaningful characters offering important learning opportunities by serving as children’s personalized, intelligent tutors.

Authors

• Sandra Calvert, Georgetown University
  Presenting Author

• Marisa M. Putnam, Georgetown University
  Non-Presenting Author

• Naomi Ruth Aguiar, University of Wisconsin Whitewater
  Non-Presenting Author

• Charlotte Wright, George Washington University
  Non-Presenting Author

• Angella Liu, Georgetown University
  Non-Presenting Author

• Evan Barba, Georgetown University
  Non-Presenting Author