Azterketa informatizatu eraginkor baten bila

Abstract

Tests are a common method for assessing knowledge. In their classic version, they consist of fixed questions created according to the experience of the test developer. In computer adaptive tests (CAT) each person will have different items depending on her/his previous answers. To achieve this behavior, each item has to be calibrated. That is, certain characteristics have to be estimated using specific processes. This paper shows how to create a CAT and the necessary steps for calibrating the items.