Incidence of measurement disturbances in the Dental Admission Quantitative Reasoning Test

This paper describes the incidence of inconsistent person-response patterns in the Quantitative Reasoning Test (QRT), which is part of the Dental Admission Test (DAT) battery. using the Rasch person analysis approach to estimate the likelihood of a response pattern resulting from only two factors (the difficulty of the items and the ability of the person), a statistical profile of inconsistent response patterns was developed for a sample of students taking the DAT. The results of two test dates, April 1987 and October 1988, were included in this study. A random sample of 1,000 persons was drawn from each of these two test administrations. For the Fall 1988 sample, the person analysis was supplemented by a questionnaire designed to check the correspondence between a student's self-reported response behavior and the statistical profile provided by the person analysis. The analysis of the two samples of response patterns indicates a high incidence (50 percent+) of atypical response patterns on the test. This has serious implications for the admissions procedure since many of the atypical response patterns invalidate the standard scores reported for those students.