Is there a “Science of Learning” and what is in it?

A reoccurring theme (e.g., here, here, and here) of this blog is that we can improve education by leveraging findings from the science of learning. Most people in the field seem to agree with this statement, but it’s not uncommon to find people who are convinced that there is no science of learning. The reasons given are usually that 1) humans are too complex for us to be able to detect patterns, 2) that “real” learning cannot be measured by any instrument, and 3) for every piece of research that seems to “prove” one thing you can find another that says the exact opposite. People are especially likely to reject the learning sciences when the findings conflict with their extreme or half-baked views about the nature of learning and/or what constitutes effective teaching.

One of the easiest ways the uninitiated can begin their journey into the science of learning is by reading one of the many research reviews written for practitioners. Besides the ones I’ve listed below, Tom Sherrington has a good list here, and Adam Boxer has one here.

Teaching the Science of Learning (Weinstein et al., 2018)
Great Teaching Toolkit (Coe et al., 2020)
Improving Students’ Learning with Effective Learning Techniques (Dunlosky et al., 2013)
Principles of Instruction (Rosenshine, 2012)
Cognitive Architecture and Instructional Design (Sweller et al., 2019)
The Science of Learning (Deans for Impact, 2015)
Learning: What is it and how might we catalyse it? (McCrea, 2019)
Organizing Instruction and Study to Improve Student Learning (IES, 2007)
Cognitive Science Approaches in the Classroom (EEF, 2021)
Learning about Learning (NCTQ, 2016)
Applying Science of Learning in Education (Benassi et al., 2014)

For my own professional development, I thought I’d re-read each of these 11 research reviews and check how much overlap existed between them. If there really is “no such thing as the science of learning” because human complexity leads to random findings, we’d expect to find few patterns and very little overlap between the reviews. However, if there are common threads between the reviews, we should consider it our professional responsibility to investigate the implications of this information for our schools, while being mindful of the costs and boundary conditions of each recommendation.

Behold my table!

This table, which took me a couple of hours to make, reveals that there is quite a bit of overlap in the recommendations between the 11 selected reviews. It should be noted that my review wasn’t systematic or scientific: With a beer in hand, I simply started with the six recommendations in Weinstein et al. (2018) and then worked my way towards the right. I added more rows to the table as I discovered that the other papers included language around modeling, managing cognitive load, self-explanation and feedback.

While there is certainly more to teaching than what is listed in the table, if you’re like me you’ve probably never had the opportunity to really dive deep into these topics at your school. Instead, my experience has been sitting through opinion-based workshops on nebulous topics such as “How to care more about children” and “Why the way we teach is old-fashioned” and “Why one size doesn’t fit all.” When I present at education conferences, I am often the only person sharing research from the science of learning, and it’s equally clear that terms like “retrieval practice” and “interleaving” are new to many teachers in many schools.

The good news is that this list of 9 recommendations from the science of learning makes for a reasonably lean and manageable PD agenda, but only if we prioritize it.

– Zach Groshell @mrzachg