I recently attended a two-day online ELearning Guild Science of Learning Summit presented by eight L&D industry leaders. The sessions, based on solid industry research, were really informative and validated how we design and develop learning solutions for our clients at Limestone Learning. Some of the information I already knew (it was good to hear it again) and some was new (based on new research). I thought I’d share some of the highlights of the summit—ideas and approaches that I found interesting and thought-provoking.
 
The first rote memory research study was conducted in 1885 when Hermann Ebbinghaus tried to force people to learn information that was meaningless. A later study (Bartlett) proved Ebbinghaus wrong by showing people meaningful things and then researching what they recalled. He found that people remembered things that were significant to them. Another research study showed that when people were asked to translate words into stories, memory improved seven-fold (Gordon Bower & Michael Clark 1969). Repeating information (memorization) doesn’t actually strengthen our memory. To retain information in long-term memory, our working memory requires things like chunking content logically (information-mapping techniques are very useful for chunking and sequencing), priming (reactivating old knowledge before we learn new knowledge), connecting patterns and contextualized retrieval (e.g., scenarios, stories etc.—using information as we would in a real environment).

While working memory is limited to processing small amounts of information at a time, long-term memory appears to have an unlimited capacity. In long-term memory, related information is structured into “schemas,” which help us organize the information. While it’s easy to overwhelm the working memory, making it more difficult for information to get processed and stored for long-term retention and use, we can use several techniques to reduce cognitive load:

• Rather than changing format, add additional media and formats (“plus-one” thinking). For example, add a relevant video or animation to content that includes written text. Learners may choose one or use both, reinforcing the learning and having the opportunity to refer back to the static content.

• Keep it simple. Complex wording leads to reduced mental processing. Readability should typically be at a Grade 8 to 9 level, and most learners can comfortably read approximately two to three grades below their highest grade/education.

• Build microlearning to teach or reinforce content.

• Venture beyond multiple-choice questions. Ask learners to come up with responses, rather than choosing from a list of options, or—even better—think of examples where new learning would apply on the job.

• Cut out the “bling,” such as decorative images or animations, complicated navigation, extraneous “nice-to-have” information and other design features that make learners work hard just to identify and access the essential information—this adds to the cognitive load.

Strengthening learning so it’s retained in long-term memory can’t happen all at one time. The brain needs time to absorb new information. We need to pace learning so it can be remembered longer, but how much spacing and attention do learners need? It depends on how complex the new information is and how frequently it is seen/done on the job. The more complex or longer the frequency, the more practice is needed to become automated below our conscious thinking. To help keep the information in long-term memory, we can reactivate it by re-contextualizing it with a challenge (scenarios, simulations, multiple-choice questions, case studies) to help learners retrieve the information.

We use models to explain the world—our brains will automatically create models. To get learners to make complex, better decisions (to get the performance we want), a causal model is best (if I do this ... this will happen; if I do this instead … this other thing will happen). Instructional designers are bringing the wrong models into learning by telling learners the answers without the learner trying to figure out the solution, making reinforcement of learning too simple. Instead we need to give learners the opportunity to determine what the wrong model is—and to figure this out in the training session rather than on the job. A good example supporting a causal model is having learners complete a branching scenario that shows the consequences of the learner’s decisions. It’s more effective and strengthens learning because it plays out the situation and shows it going awry in context so learners recognize their mistakes as they go along. Adding a way for learners to  see their progress track in the scenario path is also very effective for learning (e.g., show a map at the bottom of the scenario to indicate where they are in the scenario and where they figured out where they went wrong).

Research has shown that instructor-led training (ILT) and elearning are not as effective as performance support tools on the job, which are many times more important to employees than formal learning.

• For employees who don’t care/know, design experiences that simulate real challenges (e.g., simulations/scenarios of situations that they would encounter—create stories).

• For employees who care (e.g., new employees), provide resources that will help them (e.g., short videos, checklists, short procedural documents) and store them in a way (e.g., corporate intranet) to make it available as needed.

Proponents of teaching to students’ “learning styles” argue that people learn better when instructional materials are presented in a format that meshes with their learning style. Unfortunately for supporters of this notion, and despite many attempts to prove its validity, it’s not supported by a significant body of evidence-based research. So why do “learning styles” persist? It’s an ingrained belief that’s hard to break and there’s big money in it—companies selling tools and books that support learning styles. Results support what many excellent facilitators already know: tailoring the instructional approach to the material produces the best results for all learners, regardless of their stated learning preferences.

For more information on the science of learning, have a read through the following:

• Clark Quinn, Engaging Learning: Designing e-Learning Simulation Games book

• Julie Dirksen, Design for How People Learn book

• Richard E. Mayer & Ruth C. Clark, eLearning and the Science of Instruction book

• Peter Brown, Henry L. Roediger III & Mark A. McDaniel, Make it Stick: The Science of Successful Learning book

• Nick Shackleton-Jones, How People Learn book

• Nick Shackleton-Jones, User-Centred Learning Design – Using the 5Di Model article

• Karl Kapp, articles on good examples of gamification

• Pam Hogle, Research Discredits Learning Styles article

• Jane Bozarth, The Truth About Teaching to Learning Styles, and What to Do Instead report

• Connie Malamed, Six Strategies You May Not Be Using To Reduce Cognitive Load article

• Connie Malamed, Nuts and Bolts: Give the Learner a Fighting Chance article

Are you using techniques and approaches that support the science of learning? If so, let us know in the comments below!