Inferential Errors in Taxometric Analyses of Ordered Three-Class Constructs

Hermann Rorschach's theory of perception is first presented, and then a method is described which was aimed at testing the theory's implications. Group Rorschach contents were selected, and two types of scoring procedures were worked out. In one case, a simple percentage of total responses was used, and in a second case a weighting was made according to whether or not the S was reporting an interpretation or a perception of the inkblot. Three all-male samples were studied, attending two different high schools. Both social class and racial differences were represented in the samples. Where one or the other of the two Rorschach scoring methods resulted in a significant correlation with independent measures of personality within the samples, it was the measure taking into consideration Rorschach's theory which reflected the over-all advantage.