How to Use Cumulative Testing to Enhance Learning Outcomes
by Kevin Patton
One of the most effective enhancements I’ve ever made to my human anatomy & physiology course was switching to cumulative testing. What I mean by that is instead of testing on each topic once, then moving on to a test on the next topic, I started testing my students on all the covered topics (thus far in the course) in each successive test.
I’ve always had a comprehensive exam at the end of the course—and eventually added a comprehensive midterm exam, too. I found that adding that midterm helped my students relearn what they’d forgotten during the first half of the semester—making them better prepared for the comprehensive final. But not a whole lot better.
As I got older and wiser—or at least grayer—and got more serious about seeking out solid research on how people actually learn new information and retain it for the long term, I realized that my thinking was sort of alongside the right track. But not fully on the right track. It finally dawned on me that I could not expect my students to really “get it” and “keep it” unless they were repeatedly challenged with a variety of test items that required them to dig back into their memories and drag out those “old” ideas from early in the course.
Learning experts sometimes call this retrieval practice. The students practice retrieving their stored knowledge and skills. One of the key elements of using retrieval practice in learning is that it is most effective when it is spaced out over time. That is, it occurs after the brain has had time to do some forgetting.
The “re-learning” and “re-remembering” that must happen after a spaced interval is one of the keys to getting it all solidly embedded into our memory. As my tai chi teacher always tells me, “you can’t master it until you’ve forgotten it.” The forgetting, making mistakes, and relearning also enhances our ability to get those concepts and skills back out of memory—thus enabling us to retrieve it when we need to apply it.
Of course, most of the effort in getting my course on the right track in this regard was getting over that same old, often insurmountable, hurdle of taking a step outside of the “way we’ve always done it.” This nearly universal mindset not only holds me back from trying new things—it encourages my colleagues and students to tell me how wrong I am when I do.
After at least a year of self-doubt, I just forged ahead and tried it. Every one of my tests now includes test items from all previous topics. I told my students ahead of time why I was doing it and why. And guess what? They were okay with it! I didn’t tell any of my colleagues what I was doing because, er, my internal voice was already telling me I was doing it all wrong.
And you know what? Without changing much else in my course that semester, the comprehensive exam grades—and even the course grades—went up almost a whole letter grade on average. In other words, my course activities and testing covered the same content, at the same level of rigor, but my students were apparent much more successful in their ability to recall the information and skills they need to solve problems at the end of the course.
This was about ten or so years ago, and I probably still have the numbers somewhere. I didn’t do a statistical analysis and I didn’t have a control group—unless you count sections of the course in previous semesters. But I didn’t—and still don’t—feel I really need that. My student grades that semester (and ever since) show a dramatic increase that I’m not willing to reverse.
Looking back through the lens of 20/20 hindsight, I can see that this should have been plain to me all along. How can we expect anyone to learn something deeply and for the long term, if they only get one chance to have their knowledge and skills challenged? Only through repeated challenges can we master concepts at a level of usefulness.
We expect our students to build a complete enough conceptual framework to see patterns and understand relationships among concepts. To really see the big picture. But do we give them enough practice to do that effectively? Or do we let them forget what they know and fail to give them those critical opportunities to relearn, thereby solidifying, key concepts?
In my experience, cumulative testing is a valuable strategy to enhance learning in our courses.
What can we use from this in teaching undergraduates?
- Adding items to every test that review all previous topics convert your testing strategy to cumulative testing. This may provide the repetitive, spaced retrieval practice that students need for learning for the long term.
- Consider using a cumulative strategy for other forms of retrieval practice in your course. For example:
- Clicker questions
- Pre-lecture and pre-lab videos or reviews
- Adaptive learning assignments
- Homework/review assignments
- In-class, small group reviews
- Recommended study strategies for individuals and study groups
Want to know more?
The critical role of retrieval practice in long-term retention
- Roediger H Butler A. Trends in Cognitive Sciences. 2011 vol: 15 (1) pp: 20-27. DOI: 10.1016/j.tics.2010.09.003
Make It Stick
- Peter C. Brown, Henry L. Roediger, Mark A. McDaniel. Harvard University Press, Apr 14, 2014. 313 pages.
This is book written for the average teacher or student to help them understand what we now know about effective learning that may be different then the traditional approaches.You really need to read this book! It’s well written, engaging, and has a wealth of great ideas.
This was originally posted here and is used with permission.