top of page
  • Writer's pictureKristin Phillips

Research About How We Learn Does Not Match How We Teach

There is some very interesting research in cognitive psychology that has been mostly ignored in education. Some of this research has been replicated many times over the last 100 years! However, some of the research doesn’t seem to make intuitive sense to teachers and, quite honestly, it is difficult to wrap your head around how you could design your curriculum to fit the research. But, not impossible.

Massed vs. Distributed Learning

For more than 100 years psychologists have observed this phenomenon over and over and it is exciting because it doesn’t intuitively make sense. The finding is that if you want people to learn something new and then retain the information longer, they do better when they learn something distributed over time rather than learn it all at once. For example, I decide to teach a bunch of people something new and then have them practice it 20 times with feedback after each practice question. One group is going to do all 20 practice questions in one sitting (massed) and one group is going to do 10 practice questions in one sitting and then wait a week to do the next 10 practice questions (distributed). I test each group after their last practice session. Both groups will fare about the same. But, when I test both groups 4 weeks later, the distributed practice group will have retained significantly more (like 30% more) than the massed practice group.


Interleaving is the practice of learning more than one concept at a time. It goes along with distributed learning. Let’s suppose you need to learn ABC and D. You could learn and practice them in either of these two ways: Blocked learning: AAABBBCCCDDD Interleaving: ABCDABCDABCDABCD Blocked learning ends to happen when we use a “unit” model of learning. Interleaving is more like the workshop approach. What research shows is that significantly more learning is retained with interleaving than with blocked learning. Although testing immediately after an AAA study period shows better results than testing immediately after the first ABCD study period, testing at the end of the entire study period shows greater retention of the concepts that were taught using interleaving. Not only that, but when the concepts taught are similar in nature, such as how to solve similar but different types of math problems, students who studied using interleaving make fewer errors in mixing up the different kinds of problems, probably because they had to learn to differentiate between ideas as they learned.


General logic might make us think that if you want to learn something really, really well, you should practice it lots and lots. This is called overlearning. Suppose you teach a group of people something new and after 5 practice trials they’ve demonstrated that they’ve got it. Common sense would dictate that if they then practice the new learning 20 times they will consolidate and really get it. However, studies show that not to be true. If you test two groups—one that used overlearning (extra practice after demonstration of understanding) and one that just demonstrated understanding after the 5 practice trials, they perform exactly the same on the test.

If we think about how we have traditionally designed curriculum, we tend to design it in opposition to the research. We teach new concepts one at a time, have students practice it over and over again, and then don’t address it again until the following year. Then, we bemoan the fact that last year’s teacher didn’t teach the kids anything!

What if we designed curriculum to be recursive or cyclical throughout a year? How could you return to topics many times throughout the year so that students had multiple opportunities to master the most important concepts? Which concepts in your subject area are the big ones? Which concepts are inter-related and would be best taught simultaneously in order to take advantage of interleaving?

More next week….

6 views0 comments


bottom of page