An experimental investigation of judges' ratings of schizophrenics' and non-schizophrenics' paintings.

Summary Four categories of paintings (N = 200), done by schizophrenic patients, art students, and education students were rated by judges (N = 13) as to whether the painter was schizophrenic or normal. There was significant agreement in correctly identifying paintings, but the percent correct, greater than chance, was only 9.99%. Judges did best on paintings with “human content.” Intelligence scores were related to correct identification of two paintings, while art experience related to proportion was judged schizophrenic for all four paintings. Stepwise regression analysis suggested that intelligence scores were artifactually related to painting scores but art experience was not.