Scoring Candidates in the Adaptive Test

Abstract

In this paper, we present a methodology able to make an automatic and rapid evaluation of the candidates in adaptive tests. The objective is to evaluate candidate level related to specific skills using minimum number of asked questions. Our proposed approach is based on quadratic algorithm in terms of compatibility. We begin this work with an introduction to the adaptive tests where we present its limitations. The second part is about main algorithms used in this work. In the third part, we present our approach of relevant selecting items. The last part is about results and its analyses.