ASSESSING THE CONSTRUCT VALIDITY OF THE ERROR CHOICE TEST: A NEW MEASURE OF MENTAL ILLNESS STIGMA Michaels, Patrick The stigma of mental illness has been well documented in the United States and remains a worldwide concern. A common methodological necessity is to accurately assess attitudes and behaviors without the adverse impact of social desirability. An inexpensive method is to develop and administer an error choice (EC) test to measure opinions. An EC test is presented as a knowledge test, but actually measures prejudice. The current study evaluated the construct validity and reliability of a new EC test on mental illness stigma by administering questionnaires to assess cognitive, affective, behavioral, and knowledge domains in a community sample (N = 203). Results indicated that EC test total score was not related to demographic characteristics, but was related to cognitive, behavioral, and affective domains. The EC test only accounted for a significant amount of unique variance when predicting appropriate personal goals for people with mental illness, R2 change = .054, F(1, 200) = 14.96, p < .001. Both the Attribution Questionnaire and EC test were significant predictors of the personal goals total score, β = .433, p < .001, β = .235, p < .001, respectively. In terms of implications, further evaluation of the EC test would be beneficial given that it is inexpensive, easy to administer, and averts social desirability while discretely measuring prejudice. The use of an EC test could prove to be a useful adjunct to the use of direct attitudinal assessment tools in mental illness stigma studies. M.S. in Psychology, December 2011 unknown 2011-11-16 2011-12 Thesis application/pdf islandora:7419 http://hdl.handle.net/10560/2565 PSYCH / Institute of Psychology Illinois Institute of Technology en In Copyright http://rightsstatements.org/page/InC/1.0/ Restricted Access