Item-fit evaluation in biased tests: a study under Rasch model