<p>
</p>
<p>A few misconceptions here too.</p>
<p>“New question surveys,” or pretests, are used to determine the approximate dificulty of the items before they are used. By pretesting items, the goal is to assure that every year’s test is of the same difficulty. Unfortunately, life steps in, so the equating proces, using equator items from a previous form, is used to help understand any difference in difficulty or strength of test populations on this year’s test compared to a previous year.</p>
<p>Comparability studies are performed about every five years in each subject. This consists of administering a previously administered test to a groups of college students. The colleges are not chosen from random, but are representative of the types of colleges that AP students typically apply to and receive credit for their AP performance.</p>
<p>The comp study tests are graded by the readers, using the same rubrics used for the AP students, and the grades are compared to those given by the college professors. This data is compared to the performance of the AP students who took the same test. All this information is then used to adjust the score distributions on the current test form.</p>
<p>The equating set is very important in determining cut points, as equating has been shown to be a reliable way of maintaining “continuity.” However, the comp study performance, essay performance and overall raw score performance of the current testing population are all factors involved in determining valid and reliable cut points.</p>
<p>Unless a table leader has been to the score setting session, information would be second hand.</p>