[Influence of mnemonic, semantic, and conceptual factors on the performance of aphasic patients in the Token Test (author's transl)].

Three experimental modifications of the Token Test were presented to 40 aphasics with predominantly expressive disturbances and 40 control patients with diffuse brain damage. On each token a human figure was drawn (a man, woman, boy, or girl); the size, color, and form of the token were given in the instructions, or the depicted figure was given as the target either in addition to or instead of the size and form. The mere fact that a figure was depicted on the tokens did not affect the performance of the patients. However, when the figures were mentioned in the instructions, the aphasics' performance improved significantly, although it never reached the level of the controls. An analysis of the error types suggested that the poor performance of the aphasics in the Token Test is at least in part due to difficulties identifying complex stimuli when the features are separately named in the instructions.