Improve student learning and metacognitive skills with frequent tests

November 14, 2017 | Claudia Stanny

Improve student learning and metacognitive skills with frequent tests

Students and instructors are accustomed to using tests to assess learning. However, a test can also be a learning experience. The testing effect has been studied (and replicated) extensively in both laboratory research on memory and in applied studies of classroom learning. Soderstom & Bjork (2014) outline the benefits of testing, describe frequently-replicated laboratory findings on the benefits of testing, and explain how metacognition improves when students experience multiple tests on course content. Soderstrom & Bjork argue that repeated testing improves learning through direct and indirect mechanisms.

Direct benefits of testing. Out of sight, out of mind and use it or lose it are two proverbs that capture the fragility of new knowledge and memories that are not used. We remember things longer when we think about them regularly. A test requires students to remember newly-learned information. Retrieving new information from memory to answer a test question modifies our memory for the tested information and makes this information easier to recall. The benefit can extend to non-tested material if students must retrieve and evaluate non-tested course information to determine that the alternative answers in a test question cannot be the correct answer.

Indirect benefits of testing. Testing produces benefits for learning in two indirect ways. First, feedback about correct and incorrect answers gives students unambiguous information about which material they were unable to remember or explain. Students can then direct more time to this material the next time they study. Second, students can reflect on the strategies they used to study for the test and adopt better strategies if they did not score as well as they expected. Students report using more effective study strategies after they had feedback from a test (Soderstrom & Bjork, 2014). Before students took a test or saw their scores, they reported that they studied by using rote repetition (flash cards), attentive reading, and focused attention. Laboratory research indicates that these strategies entail low-level thinking and are weak study strategies. After receiving their test results, more students reported adopting better study strategies. They reported using imagery, creating semantic references for information, or generating explanations in their own words when they studied. Memory research documents that these strategies are create learning that endures.

Testing improves metacognition. Metacognition refers to our understanding about what we know, our knowledge about how we learn and remember, and our ability to use this knowledge to regulate our learning and memory. Thus, metacognition describes our knowledge about which strategies are most beneficial for learning and how we use this knowledge to make decisions when we study. For example, we may decide to select a strategy for learning new information that we know is effective, even if using it requires more effort. A common metacognitive failure is overconfidence about our preparation for a test (Kruger & Dunning, 1999). Feedback from a test alerts students to gaps in their knowledge or weaknesses in their preparation. Students can use this information to devote more study time to content and skills that need the most practice and study. Students might also learn that the strategy they used to study for the test was not very effective. Students can use this feedback to adopt more effective strategies when they study new material or restudy missed items and more accurately predict future test performance.

Faculty who require students to take multiple tests can produce large benefits for student performance. Faculty might worry that administering more tests will require too much time and create too heavy a grading burden. However, Lang (2016) proposes two strategies that enable faculty to increase the frequency of testing without demanding major changes to course structure or increasing faculty work.

Create low-stakes tests at the beginning of class to give students practice retrieving information from the previous lecture or from the readings assigned for the current lecture.

  • Online quizzes can be administered and graded in eLearning. They require no additional class time and no grading time. They do require some time to create in eLearning.
  • Clicker questions can be built into lecture slides. Clicker questions can be ungraded. The questions and subsequent lecture give students immediate feedback about the correct answer. You can deepen learning if you ask students to discuss of the question and their answers in a 5-minute pair-share activity before revealing class responses or the correct answer.
  • Give students a prompt for a short essay question that concerns content you plan to discuss that day. Writing the essays forces students to recall and reflect on content, encourages students to read material before class, and creates a context for the ensuing class discussion or lecture. Collect the answers to record class attendance. These essays need not be graded. Students will get feedback about their answers from the class lecture or discussion.

Assign minute papers during class or at the close of class. Minute papers can be assigned at any point in the class session. Some instructors use them as an end-of-class reflection.

  • Create a prompt that requires students to paraphrase or explain a concept discussed in class.
  • Ask students to apply a concept discussed that day, perhaps by proposing a solution to a relevant problem in a real-world situation.
  • Maximize the need for effortful retrieval of information from memory by asking students to write without consulting their book or their notes.

Lang suggests asking students to make two copies of their essays (minute papers should be no more than half a page of writing). You can collect one copy to review answers and discuss at the start of class the following day and use to record attendance. Students can correct errors in their answers on their copy in class when you discuss (again, no formal grading for you).


Kruger, J., & Dunning, D. (1999). Unskilled and unaware of it: How difficulties in recognizing one’s own incompetence lead to inflated self-assessments. Journal of Personality and Social Psychology, 77, 1121-1134.

Lang, J. M. (2016). Small teaching. San Francisco, CA: Jossey-Bass.

Soderstrom, N. C., & Bjork, R. A. (2014). Testing facilitates the regulation of subsequent study time. Journal of Memory and Language, 73, 99-115.