The aim of this paper is to compare the efficacy of three different item se
lection algorithms in computerized adaptive testing (CAT). These algorithms
are based as follows: thr first one is based on Item Information, the seco
nd one on Entropy, and the last algorithm is a mixture of the two previous
ones. The CAT process was simulated using an emotional adjustment item bank
. This item bank contains 28 graded items in six categories, calibrated usi
ng Samejima (1969) Graded Response Model. The initial results show that the
mixed criterium algorithm performs better than the other ones.