ATD Blog

Science of Learning 101: Why Learning Should Be Hard

Wednesday, July 26, 2017

One of the things I have learned from my study of learning science research is that using intuition to select the best learning solutions frequently fails. For example, learning styles appear very reasonable. Why wouldn’t teaching people in the manner that they prefer work best? But research shows no correlation between how people like to be taught and what works best for good retention and use of knowledge. (And in many cases, what they like is the opposite of what works best, which I’ll explain later in the article.) This is the same situation with what people like to eat and what is best for their long-term health.  

Learning research shows repeatedly that deep learning, rather than surface learning, is required to learn for application. Deep learning requires mental effort and helps us learn for long-term retention and use rather than simply for instruction or testing. Table 1 shows some of the primary differences between surface and deeper learning approaches. 


Case in point: If we are training workers to be safe with hazardous waste, a surface learning approach is likely to supply the following: 

  • definitions of terms 
  • types of hazardous waste 
  • regulations related to storage, transport, and disposal  
  • methods to keep hazardous waste from contaminating humans and the environment. 

Assessment methods, under more shallow approaches to learning, often have people recall information from the presentation. This approach might meet training requirements (if performance requirements are surface), but it would not be enough to help people retain information for long-term use on the job.  
In order to help people retain information so they can use it on the job would require a deeper approach. Here are some of the elements that a deeper approach might use: 

  • relating content to prior knowledge to make mental connections  
  • restating critical content in their own words, to help form meaningful schema 
  • application of the knowledge and skills as they would use it on the job. 

So, for example, the trainer might ask people if they have ever had accidents in their homes with materials that turned out to be hazardous (connections with prior knowledge). And people might look up materials they typically throw in the trash on the job to see if they are considered hazardous and how to dispose of them safely.  
Deep learning taxes mental effort. But in return, it yields better and longer lasting memory. This is important for most organizational training, especially when longer lasting memory is needed. 


But … Hard? 

When lifting weights, a lot of people go quickly and move onto another weight exercise, where they also move quickly. But research shows better results from going slow, concentrating on form, and doing every repetition with full control. Learning science shows something very similar. Slowing down thinking and making it somewhat harder while learning can yield extremely beneficial results for remembering and use on the job. 

Robert Bjork’s work on “desirable difficulties” helps us understand that what works for learning in the short term often fails in the long run. And the long run is what matters most, as we are not (typically) training people only to pass an end-of-instruction test or to check a box. In other words, we need people to use the instruction in their work.  

Desirable difficulties during learning are learning activities that require considerable (but desirable) effort that improves long-term remembering and application. Some of the most notable examples of desirable difficulties include: 

  • spacing learning sessions apart rather than teaching people in longer sessions 
  • using non-graded testing (retrieval practice) to help them more easily recall important learning content  
  • widely varying practice situations so they mirror work situations 
  • asking people to actively generate content, such as putting important concepts in their own words and fix or build content elements. 

What these difficulties have in common is that they help people more deeply process content. Left on their own, many people would not choose deeper (and harder) processing, which is one of the reasons that learning styles doesn’t work. Most people select easier methods and think they work fine. For example, when studying for a test, most people reread material. But building and using flashcards or doing problems works far better. 
It’s also quite understandable why trainers, content experts, and instructional designers might prefer more shallow approaches, because they are less difficult to design, participants may find them less taxing, and they can cover more content. But there is a price to pay: long-term remembering and application suffers greatly. 

Elizabeth Bjork and Robert Bjork’s work help us understand that making learning too easy leads to thinking that learning has occurred when participants quickly forget and cannot actually apply. Deeper processing is critical to making the leap from learning content to work—whether from informal learning, formal instruction, reading, media, or other learning methods. 



These URLs were available at the time I wrote this article. URLs often change so if they are not active when you use them, search for them by cutting and pasting the name of the reference and PDF. 

Robert Bjork is one of the pioneers in our field. You can find many of his articles at

Bjork, R.A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe, J. & Shimamura, A. (Eds.), Metacognition: Knowing about Knowing, 185-205. Cambridge,MA: MIT Press. 

Bjork, E. L., & Bjork, R. A. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning. Psychology and the Real World: Essays Illustrating Fundamental Contributions to Society, 56-64. 

Dempster, F.N. (1990). The spacing effect: A case study in the failure to apply the results of psychological research. American Psychologist, 43, 627-634. 

Mc Daniel, M. A., & Butler, A. C. (2011). “A contextual framework for understanding when difficulties are desirable.” In Successful Remembering and Successful Forgetting: A Festschrift in Honor of Robert A. Bjork (pp. 175-198). Taylor and Francis. 

Roediger, H.L., III, & Karpicke, J.D. (2006). “The power of testing memory: Basic research and implications for educational practice.” Perspectives on Psychological Science, 1, 181-120.

About the Author

Patti Shank, PhD, CPT, is a learning designer and analyst at Learning Peaks, an internationally recognized consulting firm that provides learning and performance consulting. She is an often-requested speaker at training and instructional technology conferences, is quoted frequently in training publications, and is the co-author of Making Sense of Online Learning, editor of TheOnline Learning Idea Book, co-editor of The E-Learning Handbook, and co-author of Essential Articulate Studio ’09.

Patti was the research director for the eLearning Guild, an award-winning contributing editor forOnline Learning Magazine, and her articles are found in eLearning Guild publications, Adobe’s Resource Center, Magna Publication’s Online Classroom, and elsewhere.

Patti completed her PhD at the University of Colorado, Denver, and her interests include interaction design, tools and technologies for interaction, the pragmatics of real world instructional design, and instructional authoring. Her research on new online learners won an EDMEDIA (2002) best research paper award. She is passionate and outspoken about the results needed from instructional design and instruction and engaged in improving instructional design practices and instructional outcomes.

Be the first to comment
Sign In to Post a Comment
Sorry! Something went wrong on our end. Please try again later.