Sources used in this document:

Gleason, Sebanc, and Hartup (2000) discussed the lack of conclusive evidence linking social structure to imaginary friend creation in children. The results of studies attempting to explain why some children have imaginary companions while others do not are inconsistent, which is the reason more research such as the present study are vital to the developmental and psychoanalytic understanding of children.

Many studies have combined the categories of role-playing (children taking on a different persona than their own or impersonating characters), personified objects, and invisible imaginary companions, making distinguishing factors difficult to identify. The present study will be conducted only in regards to invisible imaginary friends to help solidify the conclusiveness of the data, rather than ignoring the important distinctions between these related phenomena. Gleason, Sebanc, and Hartrup review the findings of previous studies, including that more than half of all imaginary friends have no identifiable trigger. Additionally, children are usually not able to explain the physical appearance of their imaginary companions. Some parents have reported that children base their imaginary friends on real people or characters from movies and stories. Imaginary companions most often appear human, but can also include animals or fantastical beings such as aliens or monsters. There is not conclusive evidence as to whether it is more common for imaginary friends to appear singly or in groups, but both are common.

Bouldin and Pratt (1999) conducted a study that took this age factor into account and examined the characteristics of preschool and school-age children with imaginary companions. They distributed questionnaires to the parents of 900 children in preschools and elementary schools and obtained a sample that included participants from 2- to 9- years of age. The study revealed significant differences between children with imaginary companions (ICs) and children without imaginary companions (NICs) regarding their birth order, level of imaginativeness, and predisposition to engage in fantasy. A significantly larger percentage (29%) of ICs than NICs (14%) were reported to be firstborn children (Bouldin, 1999). There was also a significant difference between the two groups regarding how often and how well ICs and NICs played with other children. In the NIC group, 91% of the children were reported to play often and well with other children, compared with 82% of the children in the IC group (Bouldin, 1999). Bouldin and Pratt concluded that the significant difference in the number of reported firstborns in the IC and NIC groups substantiates the view that one function of the imaginary companion may be to subdue the loneliness of a child who does not have any siblings close in age. This conclusion, however, is merely an interpretation of the data collected regarding the number of siblings and friendships children have. No questions were asked in this study that directly address the issue of why imaginary companions were created.