Over a century of research shows evidence of improved student performance when retrieval practice (also called the testing effect) is used as a method of learning and assessment in the classroom. (1) In my estimation, an efficient and effective classroom cannot operate without the use of retrieval practice. It informs everyone of what students know and don’t know. It also works to strengthen memories of content while establishing powerful context for the ability to further recognize and recall information successfully in the future. I have written pretty extensively about when, why, and how I infuse retrieval practice into my high school classroom. Check out this link to see some of my articles and my book that feature this topic.
But teachers understand it would be quite impossible to review and retrieve all information covered in class. You would spend all class implementing formative assessment strategies and have no time to actually cover new content. We must select what material is worthy of our very limited classroom time. I usually filter my retrieval opportunities in class down to these three categories:
- Material that is necessary for seeing the big picture.
- Material that relates to today’s lesson.
- Material that is commonly confused.
*More on these three categories of information here.
I’ve always wondered, though, whether content that isn’t explicitly retrieved but is related to that which is retrieved, is also strengthened. More concretely, if I ask students in my psychology course to describe one theory of motivation, will other theories of motivation also be strengthened in their memory? Those topics are obviously related. It stands to reason that thinking about one concept within a domain might just improve the ability to retrieve other similar information…right?
Wrong.
Well…at least there’s a bit of research showing evidence of the testing effect for specifically practiced content and no evidence of transfer for untested, but related, material. Drs. Glaser and Richter at the University of Würzburg designed an experiment where college students were to watch a lecture before being assigned to either a restudy or retrieval practice group two days later to review the content. (2) This was repeated a total of six times and each participant alternated being in the restudy group and retrieval practice group three times each. One week later, participants took a final test of the material covered during their six sessions.
The results?
“Follow-up tests revealed a testing effect for questions referring to reviewed content. Final test performance in questions that referred to tested contents (M = .70, SE = .020) was better than performance in questions that referred to restudied contents (M = .66, SE = .017)…”
“Thus, Hypothesis 2 that testing would especially benefit the tested content was supported. More specifically, a testing effect occurred only for content explicitly tested and no transfer effect to untested content from the same lecture occurred.”
So, as previously stated, this study shows evidence that only content specifically reviewed via retrieval practice experiences the beneficial outcomes of the testing effect. Even if the information is covered during the same lesson and is similar in topic, there will be no transfer effect for material reviewed via a simple restudy method.
As a teacher intensely interested in what leads to better outcomes for my students, this research means a lot to me. It tells me to really consider what questions and reviews I set for students to complete during class. And to understand that what I do not explicitly question, they are less likely to remember..even if it is related. It also brings to the surface a question I’ve been ruminating on for quite a while about both formative and summative assessment:
Is it fair for me to include information on a summative assessment if I have not tasked students with retrieving and using that content previously in class or for homework?
The more I learn about human memory processing and learning, the more I’m not sure it is fair. This again intensifies my focus on exactly what students are retrieving during review sessions in class. I cannot and should not assume students know information from class after simply sitting through a lesson. As counterintuitive as it may seem, a lot of forgetting happens during learning. And the material we think about and use less often tends to be the more easily forgotten information. I need my students frequently assessing their knowledge via retrieval opportunities to (1) inform both themselves and me of their level of knowledge and understanding and (2) to further establish and strengthen those memories to stave off forgetting. To a certain degree, memories operate on a ‘use it or lose it’ principle.
So, understanding the very real constraints of time in the classroom and the findings from this study, what can I do to ensure my students are thinking about what they need to think about?
A very accessible and tangible change is to alter my questions. I mostly ask very pointed questions with very straightforward answers. I appreciate this because it makes it quite clear to the students what they should know. But maybe I need to ask more questions that are goal-free in nature (I’ve written previously on the goal free effect). What I mean by that is to be a bit more open-ended. If I need to assess my students’ understanding of the process of hearing, I could ask a question like this:
Name two structures of the inner ear and describe their function.
Or, I could ask for this same information in a more goal free manner:
Tell me everything you know about how we hear, naming and describing the functions of as many structures as possible.
Now, instead of having students simply recall two structures, they are given the opportunity to recall and use many, many more. And, as the present research seems to indicate, only the information explicitly recalled will experience the improved outcomes of the testing effect. By asking more goal free questions when possible, students aren’t necessarily hampered by the constraints of very specific questioning.
Does this mean I will completely change how I ask questions in my class? Absolutely not. There are certainly times when I need my students to understand whether they properly understand very specific concepts, so very pointed questions are needed. And, more goal free questions usually require more of an investment in feedback…which can certainly be worth the time, but is something teachers must consider when setting these types of questions.
As a teacher who is constantly trying to better my craft, I love these types of studies. They have a clear objective and outcome that can easily lead to applicable changes in a large swath of classrooms. I know no study is perfect (and we certainly need more research on this topic), but they make me consider what I do in the classroom and provide concrete ways in which I can create incremental improvements in student learning. I think that’s what being a professional is all about. I think that’s what constantly trying to better yourself for your students is all about.
- Roediger III, H. L., & Karpicke, J. D. (2006). The power of testing memory: Basic research and implications for educational practice. Perspectives on psychological science, 1(3), 181-210.
- Glaser, J., & Richter, T. (2025). The Testing Effect in the Lecture Hall: Does it Transfer to Content Studied but Not Practiced?. Teaching of Psychology, 52(1), 17-26.
Feature image by Marta Wave: https://www.pexels.com/photo/black-girl-writing-on-paper-while-reading-book-6437841/
Blake,
I don’t agree with you about fairness. It’s the job of the teacher to help students apply the strategy themselves and NOT to do it for them for everything they have to remember.
paul