Evaluating Quality of English Achievement Test Items Using Rasch Analysis: A Preliminary Approach

Abstract

The Rasch model has newly been applied in educational assessment as an evaluative tool for tests’ validity. The Rasch analysis estimates the probability of a particular answer according to person’s ability and item difficulty parameters. In the present study, a sample of 190 students was chosen through systematic sampling technique. An English achievement test having 40 items was administered on high and higher secondary school students. One mark was granted for correct response while zero score was granted to incorrect answer. Respondents’ scores were evaluated and respondents scores as well as item score were enlisted in the form of a table. The results were analyzed using Rasch model. The findings of Person characteristics curve (PCC) and Item characteristics curve (ICC) demonstrate that persons and items are free from each other. Rasch calibration also highlighted three test items showing unusual response and suggested the removal of these items for effective measuring. The results revealed that Rasch model is an effective evaluative tool and provide more precise information that can be used by test developers to re-examine potentially problematic test items, eventually improving the test validity.