Social Environments and Imaginary Companions

• In Event: 1-142 - A Multicontextual Examination of Imagination in Development: Exploring Culture, Social Environment, Cognition, and Morality

Integrative Statement

Although pretend play is practically ubiquitous, only about 40% of children create imaginary companions (IC; Moriguchi & Todo, 2018). One hypothesis is that children create ICs because they are lonely or lack certain real-life relationships (Manosevitz et al., 1973; Taylor, 1999). While ICs do provide relationships (Gleason et al., 2002), this compensatory hypothesis has found minimal empirical support beyond a weak connection with being an only child/firstborn (e.g., Ames & Learned, 1946; Singer & Singer, 1990). Other studies show no link to birth order (Hurlock & Burstein, 1932), number of friends (Gleason, 2004), household members, pets, or playmates (Manosevitz et al., 1973). No study, however, has examined ICs relation to multiple components of children’s social networks simultaneously, nor has any study tested the idea that a “missing” relationship (e.g., no sibling) might predict the type of relationship a child creates with an IC rather than an IC’s presence/absence.

Parents of 83 preschool children (M = 4.43 years; SD = 0.66) with and without ICs completed surveys regarding children’s social networks, including number of siblings, close friends, extended family, and other close adults. Children with ICs (n = 42) categorized these relationships as friend, parent-child, sibling, or pet relationships.

Social network components modestly predicted IC creation. A logistic regression predicting IC creation from social network variables successfully predicted 57.3% of cases, X2(4) = 9.70, p = .046. Having more siblings and extended family members marginally decreased the likelihood of IC creation; however, number of close friends or other close adults had no effect (see Table 1). Although the proportion of only/firstborn children was greater among children with ICs (50%) than those without (40.5%), IC status did not relate to birth order, X2(2, N = 82) = 3.47, p = .176.

We found no evidence that children create ICs to compensate for relationships missing in real life. In fact, children who created imaginary friendships had marginally more real friends (M = 3.91, SD = 1.76) than did children who created other kinds of IC relationships (M = 2.73, SD = 1. 56), t(35) = 2.02, p = .051. Only children (50%) were also no more likely than those with siblings (35.7%) to create an imaginary sibling, X2(1, N = 38) = .629, p = .428. Birth order was marginally related to relationships created, X2(6, N = 38) = 12.29, p = .056, but not in a way consistent with the compensatory hypothesis. For example, only children (50%) did not create sibling ICs more often than other firstborns (55.6%; see Figure 1).

In contrast to the compensatory hypothesis, the composition of children’s social networks plays a minor role in IC creation, suggesting ICs complement rather than compensate for social realities. Children also are not using ICs to fill specific relationship gaps. If anything, experience with friendship, in particular, might provide a template for creating an imaginary friend. As with real relationships, the developmental significance of imaginary relationships likely lies not in their presence or absence, but in their quality and social affordances.

Authors

• Tracy Gleason, Wellesley College
  Presenting Author

• Mona Husain Baloch, Wellesley College
  Non-Presenting Author

• Shaye Firmin, Wellesley College
  Non-Presenting Author

• Shruti Samala, Wellesley College
  Non-Presenting Author