| Abstract: | Reports an error in the original article by De Soto and Liebowitz (Journal of Abnormal & Social Psychology, 1956Nov], 53, 334-337). Some incorrectly calculated probabilities of Type II errors were given. Recalculated probabilities are given in Table 1. These are probabilities of Type II errors for 2-tail t tests of the difference between mean vocabulary scores of high and low perceivers in case of certain hypothetical true differences in various experimental conditions and groupings, as explained in the original article. The probabilities in Table 1 are substantially greater than those given earlier, but low enough that a complete absence of significant t's would be surprising if a large true difference existed. All t's obtained were nonsignificant, and the average difference was in fact in the wrong direction. It may be noted that 1-tail t tests might have been used, since the direction of the difference was hypothesized. One-tail t tests would have had lower probabilities of Type II errors than 2-tail t tests for the hypothetical true differences listed, and would also have yielded uniformly nonsignificant results. (The following abstract of this article originally appeared in record 1958-03650-001). This experiment was a replication of that of Krech and Calvin (Levels of perceptual organization and cognition. J. abnorm. soc. Psychol., 1953, 48, 394-400) using modified techniques. The correlation previously reported by Krech and Calvin between intelligence and perception, did not obtain upon replication. The data confirmed their finding that progression over trials was usually from low to high perceptual organization, but did not show a correlation between perception and stimulus-exposure time from 0.01 to 1.00 second. (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
