The Use of Person-Fit Statistics To Analyze Placement Tests.Report as inadecuate

The Use of Person-Fit Statistics To Analyze Placement Tests. - Download this document for free, or read online. Document in PDF available to download.

Person fit is a statistical index that can be used as a direct measure to assess test accuracy by analyzing the response pattern of examinees and identifying those who misfit the testing model. This misfitting is a source of inaccuracy in estimating an individual's ability, and it decreases the expected criterion-related validity of the test being used. In placement tests, where individuals are usually classified based on their estimated ability levels, misfitting results in misclassification of individuals, which negatively affects the student as well as the academic organization. The study applied person-fit statistics as estimated by the standardized log-likelihood (l subscript z) index to analyze placement tests used by freshman college students (samples of 1,558 and 1,077). Results support the use of person-fit statistics to analyze placement tests effectively. The distribution of person-fit statistics on each test was very close to the standardized normal distribution, and both tests accurately assessed the intended student's ability. Results also show that ability level and person-fit statistics are not related. As for the aberrant response resulting from misfitting persons, two unusual common responses have been observed: missing most of the items at the end of the test and missing more easy items than expected at the beginning of the test. (Contains 2 tables and 23 references.) (SLD)

Descriptors: Ability, College Students, Foreign Countries, Goodness of Fit, Higher Education, Statistical Analysis, Student Placement, Test Use

Author: Dodeen, Hamzeh



Related documents