You administer the sample item to a population of students and construct with testing software an item strip.
(a)
Using a standard statistical analysis packages you can determine the ability level of the students that selected that answer. The obvious expectation is that the key be the answer of choice more frequently for higher ability students. Trick incorrect answers may attract the best students; these trick distracters should be avoided in a quality placement test because student performance will be negatively correlated with ability.
(b) Item strips should also contain the r-biserial value and a measure of difficulty level that is frequently called delta. The r-biserial measures the level of correlation between success on an item and success on the placement test. In quality tests this correlation is high. The delta value is used to give a proper distribution of questions on the placement test.
