The purpose of this paper is to present a hypothesis testing and estim
ation procedure, Crossing SIBTEST, for detecting crossing DIF. Crossin
g DIF exists when the difference in the probabilities of a correct ans
wer for the two examinee groups changes signs as ability level is vari
ed. In item response theory terms, crossing DIF is indicated by two cr
ossing item characteristic curves. Our new procedure, denoted as Cross
ing SIBTEST, first estimates the matching subtest score at which cross
ing occurs using least squares regression analysis. A Crossing SIBTEST
statistic then is used to test the hypothesis of crossing DIF. The pe
rformance of Crossing SIBTEST is evaluated in this study.