en.wikipedia.org/wiki/Imaginary_friend

The Brinthaupt & Dove abstract lays out the three studies clearly:

• Study 1: differences in self-talk between different age groupings.
• Study 2: children without siblings reported more self-talk than children with siblings.
• Study 3: respondents who reported having an imaginary companion in childhood reported more self-talk than those who did not.

That directly contradicts the Wikipedia sentence, which claims the first study showed that only children who create imaginary companions engage in high levels of positive self-talk and have more positive social development. The paper's abstract does not report any such Study 1 result, and it does not say that Study 1 combined only-child status with imaginary-companion status or measured "social development."

A later review summarizing Brinthaupt & Dove likewise describes the findings as higher self-talk among adult only-children and among people who reported having had an imaginary companion in childhood—again, not a first-trial finding about "positive social development."

This claim is incorrect because adolescent imaginary companions are well documented in the literature.

A PubMed-indexed 1997 paper is literally titled "Imaginary companions in adolescence: sign of a deficient or positive development" and describes hypotheses about why adolescents create imaginary companions. That alone shows the phenomenon is not "unheard of."

A later longitudinal study likewise examined middle school children with imaginary companions. Its abstract says the authors tested whether having an imaginary companion in middle school was associated with later outcomes, using a sample of 152 middle school children. The full-text results section reports a current-IC group with a mean age of 12.2 years. So adolescents with imaginary companions were not only "heard of"—they were the explicit subject of multiple peer-reviewed studies years before this text was written.