4.

How to Put Microlearning Into Action

In chapter 2, we discussed learning design, supporting our claims with applicable learning theories and domain principles. Then in chapter 3, we used our discussions of how we learn (theories), what we learn (domains), and why we learn (classification) from chapter 2 to help us select the appropriate use case for microlearning.

Now it’s time to decide how to put microlearning into action. What provides the best effect? This chapter focuses on taking that use case and incorporating it in a manner that maximizes the outcomes. And we’ll share research that can help you understand how to do just that!

To be successful, you need to make sure research and evidence is on your side when you embark on a training or learning initiative. Learning professionals understand that the right application of the right techniques at the right time is what makes a learning program work—not the technique itself. Remember that microlearning is not a panacea in terms of learning design. Without this understanding, you risk wasting organizational resources, not to mention your own time and effort. Use the research and results from the studies in this chapter to inform your implementation of microlearning and to make key decisions about when microlearning will be effective—and when it won’t.

Microlearning Research

How far back does microlearning go? The term microlearning is typically believed to date back to 2002 (Friesens and Hug 2007). However, even before microlearning entered the lexicon, many organizations, teachers, and trainers talked and wrote about chunking learning into small pieces, creating small bites of content, or some other terminology to describe the presentation of small amounts of content to learners. Researchers have studied the presentation of small pieces of content for decades, and there is evidence that the concept of microlearning traces back more than 100 years. There is a large body of research to examine as we describe the application and usefulness of microlearning.

This chapter focuses on a select group of empirical studies and documented research that was approached using sound scientific practices. Our goal was to leverage clear, supported evidence to validate the use of microlearning. Narrowing the body of knowledge to articles that take a scientific approach to microlearning meant that there were fewer resources from which we could draw, but it also removed the “noise” or aspirations for microlearning. Our intent was not to review every study on the subject; rather this chapter aims to provide scientific evidence for the use of microlearning as a tool to deliver instruction.

The Forgetting Curve

In the late 1870s, Hermann Ebbinghaus began to study human memory and, shortly thereafter, the concept of forgetting. Using himself as the subject, Ebbinghaus tirelessly and rigorously experimented with his memory. In 1885 he published a book on the subject titled Über das Gedächtnis, which was then translated into English and published as Memory. A Contribution to Experimental Psychology in 1913. His study and the publication of his findings was a huge contribution to the study of memory and humankind’s understanding of how to improve memory.

The lasting feature of his work is the Ebbinghaus forgetting curve (Figure 4-1), which portrays a plot of memory loss over time. Ebbinghaus found that memories decayed over time. However, he found that when he reintroduced content at certain prescribed intervals, he could diminish the forgetting process.

Ebbinghaus’ research has been replicated many times. Researchers Radossawljewitsch (1907) and Finkenbinder (1913) conducted studies similar to Ebbinghaus and found similar results in terms of memory and forgetting. Heller and colleagues (1991) conducted a replication study of Ebbinghaus’s work in Germany in the early 1990s. In 2015, Jaap M. J. Murre and Joeri Dros (2015), researchers at the University of Amsterdam, discovered similar results by replicating Ebbinghaus’s work as closely as possible, thus rendering the curve valid.

Figure 4-1. Ebbinghaus Forgetting Curve

Having stood the test of time (so far), the findings and conclusions of Ebbinghaus clearly indicate that while memory decays over time, that decay can be reduced by reintroducing the items to be learned at specific intervals.

When it comes to using the Ebbinghaus forgetting curve as justification for microlearning, we do have a word of caution: The experiments, even the replicated experiments mentioned here, were all conducted using nonsense syllables as the content to be “learned.” Other research has shown that words with specific meanings tend to have more durability than nonsense syllables. So, while the forgetting curve is real, it might not be as applicable to content that has a deeper meaning than nonsense syllables. When the information is more meaningful, the decay or forgetting process should occur more slowly than the Ebbinghaus forgetting curve would indicate. For example, if you needed to learn a series of acronyms at a new job, the forgetting curve would likely not be as steep because those acronyms are related to your job and thus have meaning to you.

It’s important to remember that humans do not store information by making a literal copy of that information. Instead, we learn by encoding and storing new information based on how it relates to what we already know. We map new information to current information and link our new information with existing information (Bjork 2012). The introduction of stimuli after an item has been learned will spark memory and aid with retention and recall. In fact, if combined with deeper meaning than nonsense syllables, we can safely assume the forgetting curve will be slowed even more dramatically than Ebbinghaus’s work suggests.

In addition to adding meaning, two other types of methodologies can be used to help learners recall information and both can be useful tools as you design microlearning programs. One is called the spacing effect and the other is called the testing effective.

Retrieval: Spaced, Practiced, and Changed Behavior

Much of this chapter is intended to help you determine when and how often to provide learning moments, especially when using microlearning. A little too much information may inundate, demotivate, and even desensitize the learners, and then all your effort would be lost. The research we have pulled together here highlights the best method to use given the use case selected.

As we noted when discussing the forgetting curve, microlearning is great for ensuring our memory is in top condition! So how do we know when to retrieve? Let’s look at three different ways we can benefit from retrieval. In the first classification we look at spacing, we then look at opportunities to practice, and, finally, we examine retrieval for impact on behavior.

Spacing

The spacing effect, or spaced retrieval, is an instructional concept that involves providing learners with content spaced over time and has been shown to be an effective tool for aiding retention (Carpenter and DeLosh 2005). It helps to combat learner fatigue, as well as the potential to mix up the preceding and succeeding information they are trying to learn, and typically results in more efficient learning and improved retention (Pashler et. al 2007)

A good way to understand the spacing effect is by focusing on its opposite concept, mass practice. Mass practice is when you study a large body of content all at once, like cramming for a test. Remember doing that? You studied the night before and did really well on the test, but two weeks later you can’t remember any of it.

Mass practice or cramming presents two major problems. First is that the successive and preceding content interferes with your ability to learn the new content—concepts will become jumbled and words and definitions won’t seem to align. Your brain simply cannot completely understand, memorize, or comprehend the current information before you introduce new information into your memory. The second problem with cramming is simply fatigue. We bet you’ve studied for a test or tried to learn a great deal of information in one sitting (such as a training workshop or a two-hour webinar), only to find your brain actually “hurting” from trying to absorb too much information. We’ve certainly felt it.

In short, the spacing effect is based on the fact that memory is enhanced on a delayed test when learning events are distributed in time, rather than massed in immediate succession. The effect works because the act of retrieving information is itself a potent learning event. Retrieved information, rather than being left in the same state it was in prior to being recalled, becomes more recallable in the future; furthermore, competing information associated with the same cues can become less recallable. Using our memories alters our memories (Bjork 2012).

Spaced retrieval is most effective when engaging learners with content over an extended time and when reinforcement of the content is important for learning and application. Research has shown that the greater the amount of spacing between retrieval events, the greater the potential benefit to retention (Dobson 2013). Ideally, you would let more than 24 hours pass between the learning events, but shorter times have also been found to be effective. Learners whose practices were spaced showed better retention even eight years later than those who practiced in a more concentrated time period (Clark and Mayer 2011).

An optimal schedule for spacing content would reactivate information at the exact moment before the information was about to be forgotten by the learner. Since this is virtually impossible to know, it’s best to set up a schedule to reintroduce the information. There are two types of schedules associated with the spacing effect (Dobson 2013):

• The uniform approach. The information is presented on spaced schedules with an equal amount of time set between learning events. Thus, it’s a uniformed spacing approach with a set amount of time between each encounter with the content (Vlach, Sandhofer, and Bjork 2014).

• An expanding schedule. The amount of time between learning events gets larger with every presentation of the content (Landauer and Bjork 1978; Vlach, Sandhofer, and Bjork 2014). The spacing interval becomes increasingly longer over the course of the learning period.

Unfortunately, the research does not provide a definitive answer as to the best spacing approach. Some studies indicate uniform is better; others advocate expanded. However, there is a growing body of evidence suggesting that expanding schedules might be superior for individuals or materials that are subject to rapid forgetting (Vlach, Sandhofer, and Bjork 2014). For example, University of Nevada researcher Frank Dempster (1987) found that students retained a higher number of vocabulary definitions when a term and definition were repeated approximately every five minutes, compared with consecutively repeating the same term and definition. There also appears to be a benefit to using spacing within the confines of a single instructional setting.

Recall

Spacing tells how learners interact with content; retrieval practice is a type of spacing activity. Because there are different use cases, retrieval practice may not always be part of the mix. However, if we are attempting to practice, reinforce, or remediate, we are inevitably going to need to recall something.

It turns out that retrieving information from memory can be a powerful memory modifier. In other words, recalling information (or testing yourself) modifies the memory trace in a way that increases its accessibility in the future—in other words, it makes it stronger. Quizzing or testing yourself improves learning more than some other forms of encoding, such as restudying (Dobson 2013; Roediger and Karpicke 2006). Thus, while few individuals like them, tests are actually a good way to learn and can play a key role in microlearning.

According to Henry L. Roediger III and Jeffrey D. Karpicke (2006), researchers at Washington University in St. Louis, we can trace the use of testing to aid recall back to at least the 16th century. In a paper titled “The Power of Testing Memory: Basic Research and Implications for Educational Practice,” they cite the following from Sir Francis Bacon, the English philosopher and statesman who served as both Attorney General and as Lord Chancellor of England:

Moving up a few centuries and taking a more scientific approach, Arthur I. Gates (1917) conducted memory experiments at the Psychological Laboratory of the University of California in spring 1916. When studying children from a public school in Oakland, California, and adults who participated at the Psychological Laboratory of Columbia University (Roediger and Karpicke 2006), Gates found that retention was greatly enhanced by testing. His and similar studies have since been replicated, reporting similar results for a variety of learning materials and ages across diverse experimental designs.

Behavior

The spacing effect not only improves recall, it has also been shown to influence a person’s behavior. We saw this play out in our example of diabetic patients receiving daily text messages.

In another example, a study titled “Impact on Clinical Behavior of Face-to-Face Continuing Medical Education Blended with Online Spaced Education” found that online spaced education following a live continuing medical education (CME) course significantly increased the impact of the face-to-face course on self-reported global clinical behaviors (Shaw et. al 2011). The randomized controlled trial provided post-instruction microlearning that consisted of quizzing the learner on four clinical topics. Questions were asked every eight and 16 days, based on correct and incorrect responses (spaced retrieval), but the control group wasn’t asked any questions until week 18. Both groups were then given a behavior change survey at week 18; those who received the spaced education (microlearning) reported significantly greater change in their global clinical behaviors as a result of the program.

These two examples indicate that the act of spacing information over time and reminding leaners of suggested behaviors and content can have a positive impact on a person’s behavior. Thus, when properly employed in a training session, spaced behavioral messages can be used to guide and shape a person’s behavior.

This all helps to provide perspective on whether the topic (or topics) you have in mind are well suited for microlearning. Right now, you are probably refining the idea of your microlearning project further by thinking about how you will implement it. But it’s probably making you think about the design as well. The biggest question people always have is how long the microlearning should be. Well, there is research to support that as well.

Microlearning Duration

In the definition of microlearning we presented in chapter 1, we purposely avoided assigning a duration because we did not want to artificially constrain the concept of microlearning to a length of time. However, “How long is the right length of time for microlearning?” continues to be a perpetual question we hear.

It should be noted that there is a great deal we don’t know about optimal length, such as the relationship between content complexity and length and the relationship between relevancy and the amount of time a person is willing to spend to learn meaningful content. However, there are some research studies that provide insight into the optimal time period for microlearning.

Ralph Burns (1985) conducted research related to the attention span of chemistry students by examining the effect of certain instructional factors and how they related to study recall. He looked at presentation style and order in which ideas were presented to see how they would affect the student’s recall of materials.

Here is what he found related to time:

• The learning from the instruction was the greatest during the presentation’s first five minutes, with students being able to report about 35 percent of all ideas presented.

• The impact of the presented material declined, but remained relatively constant, for the next two five-minute portions (that is, the next 15 minutes of the presentation).

• The learning impact dropped to the lowest level during the 15- to 20-minute interval.

The findings by Burns and a similar 2013 Harvard study are relatively consistent with other studies related to duration and retention. Philip Guo, an assistant professor of computer science at the University of Rochester, also found similar results when researching human-computer interactions and online education. He analyzed 6.9 million video-watching sessions in four edX courses, and reported results related to video usage. He found that the optimal length was six minutes or shorter. He also reported that engagement times decreased as the videos become longer. For example, in videos longer than 12 minutes he found students only spent about three minutes watching the video, which means that they viewed less than a quarter of the content (Guo, Kim, and Rubin 2014).

However, Guo also found that when students were earning a certificate, they tended to engage with the videos for a longer period of time than students who were not seeking a certificate. He attributes that to motivation, because those students felt a greater need to watch the videos to eventually earn the certificate.

These recommendations from Guo’s research and the findings of others point rather clearly to a five- to six-minute segment, at least for video, of a microlearning lesson. This five-minute mark seems to be gaining consensus within the research community, supported by careful analysis and study. Mind you, this study and others like it were focused on video, and we are not suggesting that their results are representative of all microlearning types. More research is needed to see if there is an optimal overall time length for microlearning in general; we suspect the answer will be “it depends.”

Short and Sweet

Much can be gleaned by reviewing research related to the qualities of microlearning. Yes, there is still much to be learned, but you can begin implementing some of these evidence-based results into your own microlearning design and be reasonably assured you are helping the learner. The tips and concepts in this chapter will prevent you from guessing and allow you to present a solid case for why you chose to implement certain elements in your microlearning approach.

This chapter brings section 1 on the foundations of microlearning to a close. By internalizing definitions, theories, domains, use cases, and research, you put yourself and your organization on solid footing when you embark on planning and developing microlearning, the subject of section 2. This will help keep you from being swept up in the microlearning craze and bring a grounded instructional design mindset to preparing how your learners will interact with the content you provide.

Key Takeaways

Based on the findings from the various studies on the elements of microlearning, the following conclusions can be drawn:

• Mass practice or cramming is not effective because the learner can become fatigued.

• Spaced retrieval has been shown to be an effective tool for aiding retention and engaging learners over an extended time.

• Space retrieval is effective when reinforcement of the content is important for learning and application.

• The ideal time between the learning events is greater than 24 hours, but shorter times have also been found to be effective.

• Testing greatly improves retention of material.

• Five minutes appears to be a reasonable length of time for microlearning content.