Abstract: | ![]() The development of a forced-choice form of a human relations attitude test is described. After item selection (and cross-validation of item selection) of four forms, test-retest reliabilities were computed following two administrations to new samples. None of the reliabilities was greater than .43. The reasons for this low reliability are discussed. (PsycINFO Database Record (c) 2010 APA, all rights reserved) |