An Investigation of the Rasch Simple Logistic Model: Sample Free Item and Test Calibration.
MINNESOTA UNIV MINNEAPOLIS DEPT OF PSYCHOLOGY
Pagination or Media Count:
The research investigated the use of the Rasch simple logistic model in item and test calibration. Tests employing word, picture, symbol, and number analogies were administered to college students, high school students, civil service clerical employees and clients of the Minnesota Division of Vocational Rehabilitation. The results suggest that Rasch item easiness estimates are invariant with respect to the ability of the calibrating sample when an adequate sample is employed. The invariance of the Rasch item easiness estimates was shown to be related to the goodness-of-fit of the items to the Rasch model. The deletion of items with low Rasch probabilities increased the invariance of the Rasch item easiness estimates. Estimates of the amount of ability indicated by the raw scores on a test ability estimates were also shown to be invariant with respect to the ability of the calibrating sample for tests of 25 or more items, even when relatively small samples were employed. Author