May 18th, 2015

Why Open-book Tests Deserve a Place in Your Courses


Group testing

With the proliferation of learning management systems (LMS), many instructors now incorporate web-based technologies into their courses. While posting slides and readings online are common practices, the LMS can also be leveraged for testing. Purely online courses typically employ some form of web-based testing tool, but they are also useful for hybrid and face-to-face (F2F) offerings. Some instructors, however, are reluctant to embrace online testing. Their concerns can be wide ranging, but chief among them is cheating.

Of the many obstacles that web-based technologies present, combating academic dishonesty is among the most challenging. For many it is hard to envision a scenario where a student completes an online quiz (or test) without using their smartphone, tablet, or other device to look up the answers, or ‘share’ those answers with other students. Those of us who use online quizzes have experimented with lockdown browsers, randomized questions, and anything else we can find to try to ‘defeat’ the students in their quest to cheat. One potential solution is worth exploring: open-book testing.

Instead of wasting valuable time to deter cheating, open-book tests shift the onus of responsibility onto the students themselves. They are the ones who must track down answers and page through online notes. That doesn’t, however, mean we should wave the white flag. Random question generation and randomized responses are still good techniques to employ. When coupled with an open-book test, they can challenge students and reduce the relative value of cheating.

If you can’t beat ‘em, don’t try!

Cheating becomes an appealing option when the response to a question is one that can be easily Googled. A student need not read a single chapter or attend any classes, if they know their smartphone will come to their rescue. An open-book test, with challenging application questions that relate directly to the course material, can help minimize the problem. Here are some tips:

  • Draw specifically on course content/lectures. Asking students a basic identification question will send them straight to Wikipedia. Instead, ask them to analyze the author’s argument on page 34, or interpret the results shown in a diagram.
  • Keep the time tight. When time is limited students won’t be able to blindly scavenge the course notes for the answer. They will recognize the need to prepare and have some familiarity with the material or they will simply run out of time.
  • Make the questions tough. Use distractor questions that closely resemble the correct answer. Students will need more than a passing glance at the material to locate the correct response. Use application and analysis questions that challenge students to fully understand and synthesize the concepts related to the learning outcomes.
  • Recognize collaboration. The effect of randomized questions is that two students, sitting side by side, will receive different sets of questions. This ostensibly eliminates the benefit from working together. However, if we encourage students to complete the quiz with a classmate, they will find themselves navigating their notes together and collaborating to identify the correct answer. Well I hesitate to mention it, but that sounds a lot like studying!
  • Tell students you know they have access to their resources. Now it’s out in the open. It is puzzling that if students know that a test is open-book, they often assume that there is no studying required. By communicating your expectation, practicing a few questions with them (online or in-class), this tells them they need to study. Anytime I can encourage my students to interact with lecture notes, videos, and textbook chapters, it’s a win for me (learning outcomes) and a win for them (they study).

“But they aren’t learning anything that way!” you say. Aren’t they? It is true that they aren’t memorizing things and recalling them later. But that isn’t necessarily our ultimate goal. Our goal, when it comes to assessments, is to measure our student’s achievement of the course learning outcomes. If open book tests can help, why not give them a try?

Matt Farrell and Shannon Maheu are professors in The School of Language and Liberal Studies at Fanshawe College in London, Ontario.