Developmental Psychology 1977, Vol. 13, No. 3, 293-294 Sex Differences in Children's Response to Training on a Visual-Spatial Test JANE MARANTZ CONNOR, LISA A. SERBIN, ANDMAXINE SCHACKMAN State University of New York at Binghamton A brief training procedure designed to improve performance on the Children's Embedded Figures Test, a test of visual ability, resulted in significant improvement for girls but not for boys. No sex difference was observed among children receiving this training. The implications for theories of sex differences in visual-spatial ability are discussed. Male superiority in a variety of visual-spatial tasks has been widely reported (Maccoby & Jacklin, 1974). Although genetic and hormonal factors may contribute to this sex difference, it is also possible that cultural sex-role patterns which result in sexually differential childhood experi- ences may affect the development of such skills (Sherman, 1967). In the preschool classroom, for example, boys have been observed to spend more time playing with activities such as blocks and trucks, (Fagot & Patterson, 1969) which may be relevant to the development of spatial abilities. Further support for the hypothesis that females have less oppor- tunity to learn and practice spatial abilities would come from evidence that training improves fe- males' spatial abilities while having little or no effect on the abilities of males. The present study was designed to investigate the extent to which training affects sex differ- ences in visual-spatial tasks in children of differ- ent ages. The task that was used was the Chil- dren's Embedded Figures Test (CEFT) on which male superiority has been observed as early as age 7 (Witkin, Oilman, Raskin, & Karp, 1971). We developed two sets of training materials, which were designed to facilitate performance on the CEFT after a relatively brief training period. This research was supported by National Institute of Mental Health Grant 1R01 MH 25751 and a grant from the Research Foundation of the State University of New York. We wish to thank Norman J. Sweeney, superintendent, Stuart D. Neismith, and Ronald Lil- ley, principals, and the staff of the Chenango Forks School District for their assistance and cooperation. Stan Kaufman designed the training figures; Richard Burright provided thoughtful comments on an earlier draft on this manuscript. Requests for reprints, or a more complete report, should be sent to Jane M. Connor, Psychology De- partment, State University of New York, Bing- hamton, New York, 13901. One set of materials (used in the overlay condi- tion) consisted of five complex pictures with a diamond shape embedded in them. The complex figures could be successively simplified by a re- moving a series of three overlays, which removed increasing amounts of extraneous detail from the pictures. The overlays could then be repositioned so that the child could directly observe the diamond shape becoming visually transformed as more detail was added. To determine the effect of simple exposure to these training figures without the overlay procedure, a copy of the complex figures without the overlays was used for training a second group of children, the flat figures group. All children were asked to find the diamond shape in the training figures and were shown the diamond if unable to do so. A control group re- ceived no training but spent an equivalent amount of time (5-10 minutes) talking informally with the experimenter. One hundred and thirty-three children in grades 1, 3, and 5 (ages 6 through 10) were ran- domly assigned to one of the two training condi- tions or the control group. After training, all chil- dren were then tested on the CEFT as described in the instruction manual (Witkin et al., 1971). As predicted, the effect of training was not the same for males and females. Among the boys, the performance of the three groups did not differ; the means for the control, flat figures, and overlay groups were 16.09, 16.57, and 16.22,F(2, 115)< 1. Among the females, the overlay group, with a mean of 17.56, scored significantly higher than the flat figures group, with a mean of 14.74, and the control group, with a mean of 13.58, F(l, 115) = 4.23, p < .05. The latter two groups did not differ from each other, F(l, 115) < 1. Comparing males with females, there was a marginally significant tendency for males to score higher than females in the control condition, F(l, 115) = 3.52,p < .10. This tendency was reduced in the flat figures condition, and the direction of 293