Human figure drawings of gender-problem children: a comparison to sibling, psychiatric, and normal controls.

The Draw-a-Person test was used to evaluate a number of questions regarding children referred for potential problems in their gender-identity development (N = 36). Sibling (N = 31), psychiatric (N = 23), and normal (N = 30) children served as comparison groups. The major results included the following: (1) The gender-referred children were more likely to draw an opposite-sex person when requested to "draw a person" than were the other three groups; (2) the gender-referred children who drew an opposite-sex person were more likely to play with opposite-sex toys and dress-up apparel on a free-play task than were the gender-referred children who drew a same-sex person; (3) the gender-referred children drew taller opposite-sex persons than same-sex persons; (4) using Koppitz's (1968) criteria, the normal children had a smaller proportion of emotional disturbance indicators in their same-sex drawings than did the other three groups. These findings were discussed with regard to psychometric and interpretive issues in the assessment of children with atypical gender-role behavior.