Geometric figure-rotation task and face representation in dyslexia: role of spatial relations and orientation.

Two tasks tapping various kinds of spatial-relations representation were given to test the role of spatial representation in writing and reading. (1) An intentional rotation task using six geometric (Bender-Gestalt) figures was performed by 356 children (ages 8 to 15 yr.), namely, 252 normal readers and 104 dyslexic children. (2) A draw-a-person test with the face from the front was added, since previous studies showed a significant positive correlation between un- or under-developed writing and reading abilities on one hand, and the specifically distorted pictorial representation of the upper part of the human face ("neolithic faces") on the other. Compared to normal readers, the dyslexic children not only drew significantly more "neolithic faces" but also made more errors of spatial displacement (up/down or right/left) on parts of asymmetric figures, while among both groups there were similar percentages (67.06% and 65.38%, respectively) of children who made no errors in the merely global rotation of the figures. The specific difficulty in the performance of a subgroup (78.85%) of dyslexic children on the figure rotation task appears to be analogous to certain errors made with lexical signs, where the global shape is preserved, while its asymmetrical parts may be displaced: (b = d = q = p, N = Z). Analogously, the "neolithic faces," while misrepresenting (obliterating) the spatial-relationally complex and individualized bridge of the nose area, preserve merely its global configuration.