To Retain New Learning, Do the Math

Every teacher experiences the frustration. Content and skills taught throughout the year seem to abandon students during springtime standardized testing. “How can they not know this?” thinks the the teacher. “We learned this back in November.”

Recent research reveals some likely causes, and the principles for retaining new learning may not be intuitive to us as teachers. For example, multiple retrievals rather than multiple exposures promote better retention of new learning.1 In other words, the more students are required to recall new content or skills, the better their memory will be. Reviewing the material with students does not have the same effect. The students must be engaged in activity that requires them to recall the material. Even when students recall details incorrectly, if the teacher promptly provides the necessary instructive feedback, engaging students in recall of the material fosters better retention of new learning than a teacher-led review.2

But how often should teachers be engaging students in recall of newly learned material? Two findings provide answers.

First, repeated recall should occur frequently immediately following new learning. For example, a teacher who teaches students to add fractions should engage students in recall and use of that material several times over the school days immediately following instruction. Again, even if students do not recall the skill correctly, requiring recall combined with immediate instructive feedback is more effective than reviewing the skill.3

Second, once the initial period of learning and multiple retrievals is past, students still need to be engaged regularly in recall of the material. In general, students need to recall the material after a delay of 10 to 20% of the time between initial learning and final testing.4 For example, if students learn a new skill with only a month of school (about 20 school days) remaining, they should be engaged in recall of that skill every 2-4 days. This increases the likelihood that the new learning will be part of their knowledge when they begin the following school year. (Ideally, they would be recalling that skill every 7-14 days over a 10-week summer break!)

So, let’s go back to our opening scenario: a teacher teaches material in November that students need to recall for testing in May—a gap of about six months, or about 120 school days. To increase the likelihood that students will recall the material in May, they should be engaged in retrieving it every 12-24 days, once or twice a month, probably closer to every 12 days for the first few months and every 24 days for the last few months. It is critical that every retrieval be accompanied by immediate instructive feedback.

One more principle helps us design activities that engage students in retrieving new learning. The more material students are required to recall, the better. For example, if students are required to retrieve or construct an explanation of how to add fractions and actually apply the skill to add fractions, their retention will be greater than if they are merely required to apply the skill.4

According to this research, many of our classrooms may be structured for minimal memory retention. If we begin every school year reviewing material from the previous years and spend the second half of the school year introducing new material, students are less likely to retain the new learning in future school years because they were not engaged in recalling it throughout the school year. We need to be teaching more new material at the beginning of the school year and reviewing that material as the school year progresses. Perhaps this helps explain another common teacher frustration: the “They should have learned this last year” syndrome that we’ve all experienced.

Retrieval + Instructive Feedback = Retention of New Learning.

1. Devachi, L. The Limits of Memory: How to Maximize Your Memory Trace. Presented at the 2008 North American Neuroleadership Summit, New York.
2. Baddeley, A., Eysenck, M. W., & Anderson, M. C. Memory (New York: Psychology Press, 2009), p. 70-78.
3. Ibid. 74.
4. Ibid. 82.

Memory Thinking & Experience

The Architecture of Learning Instructional Design Model recognizes four cognitively-distinct processes: experience, comprehension, elaboration, application. These four represent learning’s core processes—processes that optimize each other’s contribution to learning. (A fifth process, intention, involves responding to current, “real-world” circumstances with previously learned content and/or skills.)

While reading Teaching for Wisdom, Intelligence, Creativity, and Success (Sternberg, Jarvin, & Grigorenko, 2009), I discovered intriguing parallels between Architecture of Learning’s core processes and the authors’ identification of “four types of different thinking skills: memory, analytical skills, creative skills, and practical skills” (p. 19). While not a perfect match, the similarities are worth exploring.

Sternberg, Jarvin, and Grigorenko begin with memory because “if you have no information or skills to draw upon, there is nothing to analyze, create, or apply” (p. 19). Memory thinking involves gaining the information needed for analysis, creativity, or application, and storing that data with little or no additional processing. For example, I can recite the freezing points on both the Fahrenheit and Celsius scales but not explain their significance. I can state the capital of New York State but not describe its significance. I have data, nothing more.Some verbs associated with this type of thinking include get, take in, obtain, and seek.

This description fits nicely with Architecture of Learning’s explanation of experience. During experience, the brain receives data from the senses—data that provides the raw material for additional processing and deepened learning.

What does this parallel reveal? First, a growing consensus suggests that merely obtaining data represents shallow learning. Encoding a fact simply provides data for beneficial thinking, and that thinking is what drives deeper learning.

Second, a growing consensus suggests that learning is a process—a cognitive process that requires different types of thinking at differing stages. Architecture of Learning, with its core processes and strands, represents teaching based on this view of learning. By equipping teachers with such instructional design tools, we can tailor instruction to the cognitive processes that empower learning.

The parallels do not end with Sternberg, Jarvin, and Grigorenko’s memory type of thinking and Architecture of Learning’s core process of experience. In the next posting we’ll explore the connections between analytical skills and comprehension. The links to critical thinking are especially interesting!

By the way, if you are on Twitter and would like to follow my “tweets,” I can be found @kdwashburn.

Sternberg, R. J, Jarvin, L. & Grigorenko, E. L. (2009). Teaching for wisdom, intelligence,creativity, and success. Thousand Oaks, CA: Corwin.