Aphasics (N=30) and right hemisphere brain damaged controls (N=24) were given four different Token Test displays with varying degrees of organization of the tokens: in half of the experimental conditions the organization of the display was explained to the patients, while in the other half the four different displays were presented in a random order. When the responses are dichotomized as "right" or "wrong", there is no difference between the experimental conditions; when, on the other hand, the individual dimensions (size, colour, shape) are scored separately, there is one significant difference between conditions: when aphasics were informed about the organization of the display and when the tokens were organized with respect to all three dimensions, the number of error scores for size was reduced. © 1987.
Cohen, R., Gutbrod, K., Meier, E., & Römer, P. (1987). Visual search processes in the token test performance of aphasics. Neuropsychologia, 25(6), 983–987. https://doi.org/10.1016/0028-3932(87)90103-5