Tenacious inferential illusions: Experimental fact vs. critical fancy


Requests for reprints should be sent to Herbert Mirels, Department of Psychology, Ohio State University, 164 West 19th Avenue, Columbus, Ohio 43210.


Tzeng and Tzeng's (1982) assertion that the results from two indices of estimated item coendorsement should not have been combined in Mirels' (1976) demonstration of inferential illusions is shown to be based on erroneous suppositions. Evidence for the equivalence of the two formats makes it clear that the separate presentation of data for each index would have yielded results and led to conclusions virtually identical to those for the combined data. The objection by Tzeng and Tzeng to Mirels' use of empirical conditional probabilities as a criterion against which to assess the veridicality of Implicit Personality Theory is also shown to be without empirical foundation. In the light of available findings, the existence of striking inaccuracies in the culturally prevalent Implicit Personality Theory seems well established.