Some Methodological Issues with “Draw a Scientist Tests” among Young Children

Abstract

Children’s stereotypes about scientists have been postulated to affect student science identity and interest in science. Findings from prior studies using “Draw a Scientist Test” methods suggest that students see scientists as largely white, often unattractive, men; one consequence may be that girls and minority students feel a science career is “not like me”. However, a major shortcoming in prior research is that scholars have asked children to draw only scientists, thus making interpretations of earlier research findings ambiguous. We added other professionals to compare how 616 drawings of teachers, scientists, and veterinarians by 206 elementary school children varied by student gender, ethnicity, and grade. Students made clear distinctions: drawing teachers as most attractive and largely female, and scientists as most often male and least attractive. Aspects of the drawings suggest that scientists do have an “image problem” among children. However, large sex differences in the drawings and often‐unrecognizable gender figures in boys’ pictures lead us to question use of the “Draw a Scientist Test” as a projective test among young children.

Notes

1. Very recently, many U.S. television crime dramas now feature female forensic scientists. Their portrayal appears comparable with their male counterparts; while physically attractive and bright, these women also seem relatively reserved, meticulous, obsessed with their work, and “wedded” to their laboratories, with relatively scanty personal lives.

2. Parental consent forms were distributed to the parents of 252 students in the total school population of first, third and fifth graders. Unreturned forms and student absences reduced the number of children to 206 (82%). There were no differences by gender, ethnicity, or grade in participants versus non‐participants.

3. Only two out of three comparisons for the paired t‐tests can be used (although all three are shown for information purposes) because the third comparison is linearly dependent upon the first two.

4. Thirty‐four out of 206 children wrote captions on drawings of teachers, 41 wrote captions on scientist drawings, and 50 students wrote captions on drawings of veterinarians.

5. These results may be obtained from the first author upon request.

6. Data about unidentifiable gender in children’s drawings pertain to the second round of coding the figures. These estimates were much higher in the earlier, unreliable coding for gender in the drawings.