Genitalia in Children's Drawings: An Association With Sexual Abuse

Abstract

Many sexual abuse victims have been observed to draw genitalia on human figures. To test the hypothesis that sexually abused children draw genitalia on human figures more often than do nonabused children, drawings from 57 children, 3 through 7 years of age, who were referred to child protective services as alleged sexual abuse victims, were compared with drawings from an age-, sex-, race-, and socioeconomically matched group of 55 nonabused children receiving well-child care in medical settings. A standardized procedure to obtain drawings was followed by a structured interview to collect demographic, past medical, and developmental information. Five evaluators unaware of the children's backgrounds independently examined drawings for the presence or absence of five body parts; there was 94% agreement for all body parts and 93% agreement for genitalia. Eight children were excluded from the analysis because they only scribbled (n = 5) or because evaluators could not agree on whether genitalia were present in their drawings (n = 3). Ten percent (5/52) of the alleged sexual abuse victims and 2% (1/52) of the comparison children drew genitalia (P = .10, one-tailed Fisher exact test). The estimated relative risk was 5.4; that is, alleged sexual abuse victims were 5.4 times more likely to draw genitalia than were comparison children. Children known to have been sexually abused were 6.8 times more likely to draw genitalia than were comparison children (P = .07, one-tailed Fisher exact test). It must be cautioned that, although the presence of genitalia in a child's drawing should alert one to consider the possibility of sexual abuse, it does not prove it, just as the absence of genitalia does not exclude abuse. The drawing of genitalia should sensitize providers and influence the effort directed toward exploring the possibility of sexual abuse.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 69 citations exist upstream.