A Computer Program for Equating Two Rasch-Calibrated Tests

Abstract

EDUCATIONAL AND PSYCHOLOGICAL MEASUREMENT 1981, 41 The computer program (LINK) described calculates item difficulties for common or paired test items on two Rasch calibrated tests. In addition, it re-calibrates the test item difficulties for all nonpaired test items and centers all items onto a common scale.A voluminous number of articles has recently appeared regarding test equating with the Rasch model (Engelhard, 1980; Guskey, 1979;Lenke and Canner, 1980). Reckase (1979) found that the major axis linking technique for Rasch item difficulties did well regardless of the number of items in common between tests if sample sizes of 300 or more were used. The major axis method presented by Reckase and a similar method suggested by Wright (1979) assume that the correlation between common items shared by two tests is 1.00. Further, centering the item difficulties onto a common scale may be accomplished by a linear transformation. The purpose of this paper was to describe a computer program entitled LINK which utilizes the computational strategy suggested by Wright.Like the major axis method, Wright's procedure computes the average of the differences between the common item difficulty estimates: in which d, represents item difficulties for Test 1, d2 indicates item difficulties for Test 2, and p is the number of paired items.