This is an archived article.

December 9, 2004

Imaginary friends: Most kids have one (or more)

Imagination is alive and thriving in the minds of America’s school-age children.

It is so prevalent that 65 percent of children report that, by the age of 7, they have had an imaginary companion at some point in their lives, according to a new study by UW and University of Oregon psychologists.

The research also indicates that having an imaginary companion is at least as common among school-age children as it is among preschoolers. Thirty-one percent of the school-age youngsters were playing with an imaginary companion when they were asked about such activity, compared with 28 percent of preschoolers.

“This finding is fascinating in that it goes against so many theories of middle childhood, such as those proposed by Sigmund Freud and Jean Piaget. Having an imaginary companion is normal for school-age children,” said Stephanie Carlson, a UW assistant psychology professor.

Marjorie Taylor, a psychology professor at the University of Oregon, and Carlson are the lead authors of the study published in the current issue of the journal Developmental Psychology.

Having an imaginary companion appears to be an ongoing and changing process because a child doesn’t necessarily play with the same imaginary companion throughout childhood. Carlson said some children reported having multiple and serial imaginary companions. The number of imaginary companions described by children ranged from one to 13 different entities.

“It is somewhat of a revolving door. Children are nimble in coming up with these imaginary companions and sometimes we have a hard time keeping up with all of the ones a child has,” she said.

The researchers originally recruited 152 preschoolers, ages 3 and 4, and their parents several years ago. Each child and parents were interviewed separately about imaginary companions. The researchers also collected data on the children’s verbal ability and gave them a series of standardized tasks to assess development, or what psychologists call theory of mind. Three years later, 100 of those children (50 girls and 50 boys) and their parents volunteered for the newly published study. The children and their parents again were interviewed separately about imaginary companions. Parents also filled out a questionnaire about their child’s personality and the children took a series of standardized tasks that measured social understanding.

Children were considered to have imaginary companions if they said they had one and provided a description of it. If the companion was a doll or stuffed animal, children also had to include psychological details (such as “She is nice to me”) for it to be considered an imaginary friend.

Imaginary companions described by the children came in a fantastic variety of guises, including invisible boys and girls, a squirrel, a panther, a dog, a seven-inch-tall elephant and a “100-year-old” GI Joe doll. While 52 percent of the imaginary companions that preschoolers played with were based on props such as special toys, 67 percent of those created by school-age children were invisible, according to Carlson.

The study also showed that:


  • While preschool girls were more likely to have an imaginary companion, by age 7 boys were just as likely as girls to have one.
  • 27 percent of the children described an imaginary friend that their parents did not know about.
  • 57 percent of the imaginary companions of school-age youngsters were humans and 41 percent were animals. One companion was a human capable of transforming herself into any animal the child wanted.
  • Not all imaginary companions are friendly. A number were quite uncontrollable and some were a nuisance.

The researchers also were curious to know why children stop playing with imaginary friends.

“Imaginary companions are treated by children much in the same way as when they lose interest in toys or other activities,” said Carlson. “In many cases they simply go away, or children don’t remember. Other times children replace an old imaginary companion with a new one, or they go on to friendships with real kids to meet some of the same needs.”

The researchers also looked at childhood impersonation — pretending to be an imaginary character — and found it to be almost universal. Virtually all preschoolers pretended to be an animal or another person and 95 percent of the school-age children engaged in impersonation. The researchers did not look at impersonation in the same detail as they did imaginary companions, and were surprised that so many school-age children continued to engage in the activity. One tantalizing finding was that school-age children who did little or no impersonation scored low on emotional understanding of other people, according to Carlson.

She said that fantasy — interacting with imaginary friends and impersonation — plays a role in child development, both cognitively and emotionally. This kind of activity allows children to manage social situations in a safe context, such as practicing how to handle conflict with something that may or may not talk back to them. Cognitively it helps them deal with abstract symbols and thought, which leads them to abstract thought about their own identity.

“Imaginary companions have had a bad rap from psychologists for a long time, and there was the perception that parents were getting the message that having an imaginary companion wasn’t healthy,” she said. “But this study shows that nearly two-thirds of children have them and the striking fact is that children of all personality styles have imaginary companions.”

The University of Oregon funded the research. Co-authors of the study are former University of Oregon students Bayta Maring, Lynn Gerow and Carolyn Charley.