Tag Archive for: spacing effect

The spacing effect, if not the most studied learning factor, is certainly in the top five. As Harry Bahrick and Lynda Hall said in 2005, “The spacing effect is one of the oldest and best documented phenomena in the history of learning and memory research.”

The spacing effect is the finding that repetitions that are spaced over time produce better long-term remembering than the exact same repetitions spaced over a shorter amount of time or massed all together.

About 10 new scientific studies are carried out each year on the spacing effect (I counted 31 in the last three years). Why such frenzied dedication to exploring the spacing effect? Scientists want to know what causes it! It’s really rather fascinating!

To prepare for my upcoming conference presentation at the UK Learning Technologies conference (my presentation is now available on YouTube by clicking here), where they’ve asked me to speak on the spacing effect as it relates to mobile learning and microlearning, I’m doing another review of the scientific research. Here, on my blog, I’ll share tidbits of my translational research effort, especially when I find articles that are particularly interesting or informative.

In this post, I’m looking at Geoffrey Maddox’s review of the spacing research, which was published just last year in 2016 and is the most recent review available (although given the interest in spacing, there are many reviews in the scientific literature). He does a spectacular job making sense of the many strands of research.

In it, he finds that there are six main theories for why the spacing effect occurs. I’ve simplified his list into five theoretical explanations and I’ve ignored his somewhat jargony labels to help normal folks like me and you grok the meaning.

Five Theoretical Explanations for the Spacing Effect

  1. Spacing Prompts More Attention
    Learners may exert more attention to spaced items (compared with massed items).
  2. Spacing Prompts Retrieval
    Learners may be forced to retrieve spaced items (compared with massed items that need no retrieval—because they are still top of mind).
  3. Spacing Prompts More-Difficult Retrieval
    Learners may be prompted to engage in more difficult retrieval of spaced items (compared with massed items) and longer-spaced items (compared with shorter-spaced items).
  4. Spacing Involves More Contextual Variability
    Learners may create more retrieval routes (or more varied retrieval routes) when prompted with spaced items (compared with massed items).
  5. Spacing Prompts Retrieval and Variability
    Spacing benefits learners through both retrieval and variability, but variability, because it induces weaker traces may lead to more retrieval failure, thus lowering retrieval rates when spaced intervals are too long.

Maddox concludes that the only one that even comes close to explaining all the phenomenon in the scientific literature on spacing is the last one, which is really a combination of #2 and #4.

Count me as a skeptic. I just think we may be asking too much to push ourselves toward a unifying theory of spacing at this point. While a ton of research has been done, there are so many aspects to spacing and so much more authentically realistic research left to do that we ought to hold off on tying a pretty bow around one theory or another.

Evidence supporting my skepticism was found in the very next article that I read, where Metcalfe and Xu found more mind-wandering during massed practice than during spaced practice. This would fall into theory #1 above, not 5 (as Maddox recommended) or #2 or #4, which comprise 5.

Practical Implications

This article was not focused on providing practical implications, so it’s probably too much to ask of it. Nevertheless, it does show the complexity of spacing at a cognitive level.

Also, Maddox was pretty clear in describing how robust the scientific research is in terms of the spacing effect. He wrote, Because of its robustness, the spacing effect has the potential to be applied across a variety of contexts as a way of improving learning and memory.”

He also detailed the ways that the science of spacing is so strong, including the following:

  • The spacing effect is: “observed in different animal species,”
  • “across the human lifespan”
  • “with numerous experimental manipulations”
  • “observed with educationally relevant verbal materials”
  • “observed in the classroom”
  • and observed with memory impaired populations”

We as learning professionals can conclude that the spacing effect (1) is real, (2) that it applies to all human beings, (3) that it is relevant to most situations, (4) that it is a powerful learning factor, and (5) that we ought to be utilizing it in our learning designs!

So folks, as I wrote in 2006, we ought to be figure out ways to Space Learning Over Time, using spaced repetitions, perhaps in a subscription-learning format.

Research Cited:

Bahrick, H. P., & Hall, L. K. (2005). The importance of retrieval failures to long-term retention: A metacognitive explanation of the spacing effect. Journal of Memory and Language, 52, 566-577.

Maddox, G. B. (2016). Understanding the underlying mechanism of the spacing effect in verbal learning: A case for encoding variability and study-phase retrieval. Journal of Cognitive Psychology, 28(6), 684-706.

Metcalfe, J., & Xu, J. (2016). People mind wander more during massed than spaced inductive learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 42(6), 978-984.

Thalheimer, W. (2006, February). Spacing Learning Events Over Time: What the Research Says. Available at: http://work-learning.com/catalog.html.

The spacing effect is one of the most potent learning factors there is–because it helps minimize forgetting.

Here’s a research-to-practice report on the subject, backed by over 100 research studies from scientific refereed journals, plus examples. Originally published in 2006, the recommendations are still valid today.

Click to download the research-to-practice report on spacing.   It’s a classic!


And here’s some more recent research and exploration.

(Subscription Learning—using elearning to send learners short nuggets of learning-interactions spread out over time—is a relatively new phenomenon. As such, it is subject to the same learning curve that all new technologies entail. As a learning researcher/consultant who has been heavily pushing the idea of subscription learning in keynotes and articles recently, I’m fascinated by the human side of this burgeoning field. It’s thrilling to see new subscription-learning applications
enter the learning field, but it’s even more interesting to explore the human aspect of invention and technology-dissemination.

Eric Blumthal, who along with his business partner at count5, Gordon Eilen, set out in 2005 to create a subscription-learning platform to help sales reps better use their valuable time. They’ve created something even more broadly applicable, but their story is instructive. Eric, who had years and years of experience as a sales guy—a former VP of sales—saw a need. Sales people would be pulled away from selling to learn a ton of information, but they’d soon forget it. What Eric saw was a broken model—and a waste of time and productivity—and a loss of sales! He thought—with little or no background in the science of learning—that what was needed was a “reinforcement strategy.” When he surveyed sales organizations, hardly any had a way to reinforce what was learned. They were all leaking knowledge!

What Eric and Gordon created—along with their team—was Q MINDshare, a subscription-learning reinforcement tool. But just like all innovations, Q MINDshare has morphed and improved over the years, now incorporating learning approaches that have proven to work with real employees. And they’ve expanded beyond sales people to provide value for customer service folks, field technicians, managers, supervisors, university students, and others.

Q MINDshare – How it Works

Q MINDshare’s authoring tool is designed for non-technical subject-matter experts. The tool provides many types of interactions, including presentations, questions, videos, job aids, documents, animations, etc. They also have scenario-based open-ended question responses and then ask the learners to grade themselves. Soon they will have the capability to let learners record what they are going to say. Nuggets can be tagged with variables—for example different learning goals—and information on the learner’s progress toward these variables can be displayed to learners and provided for work-learning professionals and stakeholders on the business side as well. As an example, a Fortune 20 client of count5 uses Q Mindshare with about a dozen variables tied to the learning objectives of their training program.

From the learner’s perspective, they’ll get a notification on their computer, or phone, or tablet, which prompts them to engage in a short nugget of learning interaction.

On a computer the notification looks like this on the bottom right of the screen:

On a smartphone or tablet, it looks like this:

While most people engage immediately, the system keeps track and feeds learners new nuggets only when they’re ready. Learners typically get some sort of question to test their understanding, and then several weeks later they get a follow-up question on the same topic. The schedule of questions depends on how well the learners do the first time they answer the question.

What Eric and Gordon have Learned

Eric, Gordon, and company have learned a ton of stuff over the years. Eric first had the idea for a reinforcement tool way back in 2002 or 2003. They went to market with their system in 2005, and have been making refinements based on real-world application ever since. The following will describe some of what they learned.

Email Push Technology Doesn’t Work That Well

One of the most important things they’ve found is that people are so desensitized to emails and are so likely to ignore them that they had to invent a more potent push-technology delivery method. Specifically, they have found that they get 10-20% engagement with email pushes, but 90-95% engagement with their embedded push technology. They call this a “Cut-Thru™ technology” and refer to it as “The Q channel.”

Eric makes sure to preach the importance of using the Q channel ONLY for the really important stuff—to the organizations that count5 works with. For example, at one client, Q MINDshare is deployed to about 5,000 folks (soon to be about 25,000), but only about a dozen employees have privileges to send messages on the channel. These gatekeepers learn that keeping the channel clear of detritus is a key to keeping it valuable for employees. Indeed some companies have begun to use Q MINDshare for strategic messaging from their senior management—so that the messages get through.

The PC is still King

When people get notifications—pushed nuggets—through their PC’s, their engagement level with those nuggets is higher compared with those who get nuggets pushed to their smartphones and tablets. However, those who get nuggets pushed to both their PC and their smartphones show the highest engagement. Eric and I surmised that the PC is the most-work-focused device, whereas smartphones and tablets are often used for non-work activities. Of course, the technology landscape is always changing so the future may bring different patterns—and different user groups may have different profiles.

Prompting Attention

While there are a number of interactive methods in use through Q MINDshare, even simple presentations of information are paid more attention because the learners know that they might be asked a question about the information later.

System Design Can Lead to Continuous Improvement of the Learning

Learners can favorite any item of content in a learning interaction. They can also rate each item using a three-star scale combined with their comments. Both of these tools not only boost learner engagement, but they also help count5 learn what works best and what needs to be improved.

In many instances, when work-learning professionals review the results for one of their Q-channel threads, they often will spontaneously create follow-up learning or create job aids on information that the learners don’t seem to be getting. This is phenomenal! It’s a system that prompts us—as work-learning professionals—to get good feedback on what we’re doing so that we can build continuous cycles of improvement into our practice. Using feedback to improve our learning products is freakin’ revolutionary, but it should be common practice.

The following graphic shows one of the reports that are available. Note how high the participation rates are and how feedback and second chance opportunities lead to better proficiency.

Avoiding Chapterization Produces better Learning

As Eric and count5 have deployed subscription learning, they’ve learned that interleaving topics produces better results than teaching one topic at a time (and never coming back to reinforce them). In the learning field, we too often use such a “chapterization” strategy, when we should be using an interleaf strategy like that which count5 has found most effective. Interestingly, the research on learning suggests the same thing, showing once again that if you’re getting good feedback on real results of your learning interventions, you may find out what works without ever looking at the research. And of course, even if you’re following research-based practices (HIGHLY RECOMMENDED!) it’s still wise to verify the research recommendation with your own content and learners.

Augmenting Traditional Training and More

Originally, count5 started out to augment traditional training with follow-up reinforcement, but now Eric is pushing the concept of using Q MINDshare to provide stand-alone threads to convey learning content. His clients have also used it for prework as well.

Creating Significant Learning (and Financial) Benefits

A huge financial services company was going through the chaos of mergers and their salespeople were not selling enough title insurance. The company provided the usual PowerPoint-dense training program and followed that up with two typical elearning programs for each of their sales reps. But in addition, they also decided to try Q MINDshare as an after-training learning reinforcement device.

Not everyone at the company was completely sold on the idea, so they decided to do what the smartest companies do. They provided Q MINDshare reinforcement thread for half the sales people, but not for the other half. Then they examined the results. Their learning folks came up with a competency test that was given to all their sales folks at the end of the Q MINDshare trial. The good news for Eric and count5 was that the sales reps that used Q MINDshare performed better on the competency test by 40%—and 80% when the easiest questions were eliminated from the analysis. More remarkable were the actual sales results. Each Q MINDshare rep sold more than $200,000 more than the non-Q MINDshare reps! And this wasn’t just a few learners. There were over 200 reps in the comparison.

But even as these numbers were giving Eric thrills, there were still questions from within the sales organization. One day, as Eric was regaling the Senior VP of sales about the results, they began focusing on the biggest sale of all—a plus-million-dollar sale by Lisa, one of the best sales people in the organization. The VP decided to call her up and ask her to come immediately to his office. When she arrived—surprised if not befuddled by the sudden call to attention—the VP asked her straight-up, did Q MINDshare make the difference in her getting her million-plus deal. She said, with some humble gravitas, that she was the one responsible for the sale—that she had long been one of the most productive sales people. She also told the story of what happened.

When she’d gone to the three-day training and then taken the standard elearning courses, she didn’t feel very confident that she could sell title insurance competently to her clients. She even made a conscious decision to avoid emphasizing it with her best clients. But then, when she began to get the Q MINDshare nuggets, her sense of competence rose. Perhaps just as importantly, after about a month of engaging the Q thread, and specifically after she had engaged a Q nugget earlier one day, she was talking on the phone to one of her biggest clients and decided to broach the idea of title insurance. The rest is the history of a million-dollar sale. As you might imagine, Eric Blumthal—sitting there in the VP’s office listening to this story—was floating with delight, seeing that years of hard work and innovation had paid off big time.

Will’s Summary

Eric and Gordon of count5 are engaged in a cycle of innovation and creativity—specifically aimed at figuring out what works and what doesn’t in subscription learning. Their instincts about what might work have always been good, but now they are incorporating the latest research on learning (and spacing and feedback in particular) into their product designs. Q MINDshare is a potent subscription-learning product, worthy of our admiration. Eric and Gordon are years ahead of the field in many respects because they‘ve done the hard work of trial-observe-improve, they’ve done it time and time again, and their work shows the results.




Hello. This is Will Thalheimer! Here to advocate for and explore the idea of SUBSCRIPTION LEARNING.


What is Subscription Learning?

Subscription learning, as its name implies, provides an intermittent stream of learning-related interactions to those who are subscribed. These learning-related interactions–called “nuggets”–can involve a great variety of learning-related events, including content presentation, diagnostics, scenario-based questions, job aids, reflection questions, assignments, discussions, etc. Nuggets are short, usually presented in less than five minutes. Nuggets are intentionally scheduled over time to support learning, often utilizing research-based findings related to the spacing effect. Learners subscribe (or are subscribed) to one or more series of learning nuggets, called “threads.” Learning threads can be predesigned, selecting nuggets based on anticipated learner needs or they can be dynamically created based on learner performance.


Why is This Important?

When most of us think about learning or elearning we think in terms of providing learners with courses–large learning blocks that require learners to sit for an hour or two at a time. While courses have a long history and certainly provide benefits, we shouldn’t be hogtied by the traditional view. Indeed, the science of learning suggests that short learning events provided over time are likely to be an even more potent method. And, in today’s accelerated world, many of our learners don’t have time to devote to long periods of learning.


What this Website Will Attempt to Do

My hope is to generate converts to the subscription-learning approach, provide vetted information of value to learning professionals, and help everyone learn from our collective successes and mistakes. I’m also open to suggestions, so feel free to email me at will {dot} thalheimer {at} work-learning.com.


In 2006, I reviewed the research on the spacing effect and published a research-to-practice report, Spacing Learning Over Time: What the Research Says…

Since then I have been buoyed by the enthusiastic response to that report and by the changes that it  engendered. More training and e-learning has been built using spacing and more and more learning software has been built that incorporates the spacing effect as an inherent part of its design. If I died today, I would at least know that I'd made a small difference in our field.

Examples Wanted

I am working on an updated version of the report to include the latest research and new examples.

If you know of any examples of the use of spacing effect, please let me know. Send me demo links or disks so that I can see for myself how the spacing effect has been used. Or, just write me an email.

Testimonials Wanted

Also, if you read the original version and want to write a short testimonial about how it changed the way you build learning, that would be awesome. Just write me an email.

One Product Example: A Cameo Appearance

Just to get your juices flowing, check out this YouTube Video produced by a company who built a product with the spacing effect in mind, Yukon Learning. Yukon has built a very nice tool to support learning using the spacing effect. The product name is Cameo and the link below will take you to the Cameo website.

You can check out Cameo at this website.