"This paper evaluates certain guidelines for writing multiple-choice test items. The analysis of the responses of 5013 subjects to 630 items from 21 university classroom achievement tests suggests that an option should not differ in terms of heterogeneous content because such error has a slight but...