Although it may be difficult to prevent cheating entirely, faculty can implement steps to reduce its impact in the student learning assessment process for online courses. The following are some practical tips to prevent or reduce cheating for online testing.
1) Purposefully Select Assessment Methods
Use online testing, particularly objective test (i.e., multiple choice, multiple answer, true/false) for lower stakes assessment of student learning. In assessing student mastery of course goals and objectives, objective tests should be only one option considered among a spectrum of methods considered. Assessments should be designed to measure different types of student learning. Objective tests measure a student’s ability to recall or organize information but other types of test questions (subjective questions) can be used to assess higher order/critical thinking skills including understanding, applying, analyzing, evaluating, and creating.
2) Mix Objective and Subjective Questions
While online testing can include objective measures faculty can also include short answer or essay questions. This type of question is more subjective in nature and may demand a deeper understanding of the subject being tested. While mixing objective and subjective type questions may not discourage or stop sharing of information, it can limit the effect on the student’s final grade.
3) Use Question Pools
Rather than using a fixed number of items that remain unchanged for each administration of the test, consider creating a question pool. Questions can be grouped by any number of criteria, including topic, subject matter, question type or difficulty of question. A pool will generate an assessment with randomized questions selected by the faculty member. Pools can be created from new questions or questions in existing tests or pools. Pools are most effective when there are large numbers of questions in one group. For example, one might have a pool of true/false questions, another of multiple choice and a third for fill in the blank. The faculty member could then create an assessment drawing a specific number of questions from each of the question-type pools. Faculty can also add new questions to pools each time the course is taught to expand the variability of questions. Conversely, older questions can be removed.
3) Randomize Questions
When creating a test in Blackboard, one test option allows faculty to randomize the selection of test questions as well as the order in which they appear. The result is that students are not likely to get the same questions in the same sequence when taking a test. This strategy can address the issue of students who take a test at the same time in order to share answers. This is also relevant if faculty allow students to repeat the test. Each time this occurs, a test will be made up of questions that are randomly selected and ordered.
4) Limit Feedback
Limit what types of feedback is displayed to students upon completion of a test. Available test options include test ‘Score’, ‘Submitted Answers’, ‘Correct Answers’, and ‘Feedback’. Providing test scores is important feedback that indicates how well students have performed and should be made available. However, through a process of elimination, students may be able to determine the correct answer for each test question if their submitted answers are identified as incorrect, or if the correct answer is provided. Students could lose the incentive to both prepare for testing or to seek out correct answers by reviewing lecture notes, assigned readings, or through group discussion after completing tests. Thus, faculty might reconsider whether to include ‘Submitted Answers’ as an option to be displayed to students. This is especially relevant if faculty have allowed students to repeat tests. Each time a test was taken, students could attempt a different answer for a test question that was previously graded as incorrect. Correct answers to all test questions could eventually be accumulated and passed on to other students, or to students of future classes.
5) Set Timer
Recognizing the fact that students taking an exam that is not proctored are free to use open book/notes, faculty may decide to use the ‘Set Timer’ feature in Blackboard. Students who adequately prepared for a test may be less likely to rely on open book/notes compared with students unprepared for testing. By setting a test with an expected completion time, unprepared students could have the most to lose as they spend time going over material, and risk not having sufficient time to respond to all the test questions.
6) Display Questions One-at-a-time
If a test has more than 5 questions, do not choose the ‘All at Once’ option for displaying all the questions on the same screen. It is quite easy for students to take a screen capture of the displayed questions and share them with other students. While students can still screen capture pages with single questions, or even type them into a document, it is more time consuming and unwieldy.
For more ideas about using online testing, check out our blog page, Best Practices for Giving Online Tests.