Abstract
Imaginary companions are an intriguing part of some children’s development and tend to be created by children between the ages of 3 and 9. Some researchers have claimed they are reason for concern, others that they are a sign of giftedness, and others have found no differences between children who engage in this type of play and those who do not. If there are any differences, researchers have wanted to know if these continue into adulthood, so it is also important to know if adult and adolescent memories about childhood imaginary companions are sufficiently reliable to use for investigating group differences. The majority of research on imaginary companion play, and group differences as a function of imaginary companion status, has been cross-sectional in nature. The only longitudinal study to date has spanned three years, involving children aged 3 to 7. After a three-year interval, some of the imaginary companions identified in this research had been already forgotten by participants. The current study investigates group differences and memories of imaginary companions over an 11-year span, in a longitudinal sample of 51 adolescents, now aged 16. The participants were interviewed about their imaginary friends at 5 ½ and again when they were 16. Participants were asked if they had ever had an imaginary friend, and if so what they could remember about it. They were also asked about their previous imaginary friend, if they had one at 5 ½. Language measures were taken when the participants were 5 ½ and again when they were 16; theory of mind was measured when they were 4 and 16, and imagination, personality and psychosocial functioning were measured when they were 16. At age 16 there were no group differences between those who had an imaginary companion at some point and those who never had an imaginary companion in any of the domains that were assessed. Furthermore, almost half (43 %) of the participants at age 16 who had an imaginary companion at 5 ½ said they did not remember having one. When the imaginary companion from age 5 ½ was described to participants, the majority had forgotten about this companion; even those who said they had had an imaginary companion were often talking about one they had created after 5 ½. It is important to note that if these participants who had forgotten having an imaginary companion had only been interviewed at age 16, they would have been included in a non-imaginary companion group for between group comparisons. In conclusion, imaginary companion play does not appear to have any lasting effect into adolescence. Moreover, retrospective reports of imaginary friends may be unreliable for investigating differences between those with and without imaginary companions.