Educational program assessment studies often use data from low-stakes tests to provide evidence of program quality. The validity of scores from such tests, however, is potentially threatened by examinee noneffort. This study investigated the extent to which one type of noneffort--rapid-guessing behavior--distorted the results from three types of commonly used program assessment designs. It was found that, for each design, a modest amount of rapid guessing had a pronounced effect on the results. In addition, motivation filtering was found to be successful in mitigating the effects caused by rapid guessing. It is suggested that measurement practitioners routinely apply motivation filtering whenever the data from low-stakes tests are used to support program decisions. (Contains 5 tables, 3 figures and 1 footnote.)