An Exploratory Study of the Draw-a-Person as a Measure of Racial Identity

Abstract

The race of the figure drawn on the Draw-A-Person Test is proposed as a measure of acceptance of racial identity. Reliability and validity of this measure were explored in a two-part study. Part 1 confirmed the hypothesis that the race of the figure drawn is related to the race of the child producing the drawing. Comparison of the figures drawn by black (n = 157) and white (n = 167) children confirmed three other hypotheses based on the voluminous literature suggesting that black children are less accepting of their racial group membership than whites. These hypotheses were: (a) whites should be more likely to draw figures that are clearly white than blacks are to draw figures that are clearly black, (b) whites should be less likely to draw blacks than blacks are to draw whites, and (c) blacks should be more likely to avoid giving clear indication of the race of the figures they draw than are whites. Inter-rater reliability for the race of the figure drawn was .79. Five weeks after the initial drawing, a subsample of 28 black children drew a second figure and performed a picture choice task designed to assess acceptance of racial identity. The race of the figure drawn showed a test-retest reliability of .53 and correlated significantly with the other measure of racial identity (r₂₇ = .39).