Date of Award

Spring 1996

Project Type


Program or Major


Degree Name

Doctor of Philosophy

First Advisor

Carolyn Mebert


Two theories about why children create imaginary companions were examined, the void-stress and the psychoanalytic theories. The relationship between children and their imaginary companions, the role of parents, and fantasy choice were also examined. To these ends, children completed measures of loneliness, moral reasoning, locus of control, stress, and self-perceived competence. If they had an imaginary friend they also completed a perceived competence scale from the point of view of their imaginary companion. Children with imaginary companions also completed two measures of the relationship between themselves and their imagined friend: relationship circles and the Sibling Inventory of Differential Experience. All children were given 3 fantasy choices. Parents completed measures of parenting stress, nurturance, and restrictiveness. Results were somewhat supportive of the psychoanalytic view but not supportive of the void-stress model. Low levels of parenting stress were related to elevated levels of perceived cognitive and athletic competence in children with imaginary friends. The data on the relationship between a child and her imaginary friend suggested that children create caring imaginary companions who feel closer to the child than she feels towards it. Results also indicated that how a child portrayed her relationship with her parents was reflected in how she portrayed her relationship with her imaginary companion. It was suggested that this might be explained within the framework of internal working models. Finally, children with imaginary friends were more likely to not make a fantasy choice after the third session during which they discussed their imaginary friend. It was suggested that this might be because after a long session during which the child discussed her imaginary friendship she was attempting to balance this by making a nonfantasy, "real world" choice. Possibilities for future research were discussed with an emphasis on including data from parents and on the continued exploration into the quality and type of relationship between a child and her invisible friend.