Interview with Prof. Stephen Kosslyn

Stephen Kosslyn, Upword's Chief Learning Advisor, has dedicated his life to the science of learning. Professor Kosslyn derived the 5 principles of effective learning*: deep processing, chunking, building associations, dual coding, and deliberate practice. In this interview, Kosslyn expands on these 5 principles and explains how to implement them.

How do we know that these 5 principles work? Are the results consistent?

Each of the five principles is based on hundreds (and in some cases thousands) of studies. In formulating the principles, I had three criteria for selecting studies: First, I only paid attention to studies that were published in first-rate journals and were done very rigorously. Second, I selected studies that produced "large effects" — I wanted principles that would produce results that matter, that would make a difference in how easily and effectively students actually learn. And third, I only used studies that were consistent with many other studies, so that it was clear there is an underlying principle at work.

Who is the audience for these principles? Is it limited to students?

The principles reflect ways that the human brain works. Thus, they apply to all humans, not just students. We all can learn, and the principles capture ways that learning occurs, which apply very generally — to all topics, to all demographics.

In regard to the modern generation, do we see a need for implementing these 5 principles in higher education?

If higher education implemented the principles, students would learn more easily and effectively. One problem is that faculty would need to change their basic orientation toward teaching in order to implement them: Most large courses today rely on lectures, where faculty read notes and show slides and students scramble to write down as much as they can. This is a very effective way to teach – you can teach 1,000 students as easily as 10. But it's a terrible way to learn.

If we know anything from the science of learning, it's that being actively engaged is key to being able to take in, retain, and later recall information or use skills. And lectures are not an effective way to lead students to be actively engaged. Yes, they can be entertained and absorbed in a lecture, but this doesn't lead them to care deeply about the material and, of the course, they don't get much (typically no) feedback about how well they in fact understand the material as they write down the lecture. In contrast, active learning—such as having students engage in a debate, do problem solving, participate in role-playing games—is vastly more effective than lectures. But to insert such activities into a previously designed course requires a fair amount of work, even when faculty understand how to do it. And many (perhaps most) faculty do not understand how to use active learning effectively, which throws up another barrier: They often just don't have the time to learn this themselves, let alone to completely revamp their courses.

What is the best way to chunk information? what are some tips on doing it properly?

When we organize information, two kinds of factors play key roles: bottom-up and top-down. Bottom-up factors are characteristics of the stimuli, of what you perceive. So, for example, things that are near each other tend to be grouped together. You see "xxxx xxxx" as two groups, but the same number of x's rearranged like this "xx xx xx xx" as four groups. These sorts of factors operate both in space (as in this example) and in time (things "near each other" appear contiguously in the sequence). Top-down factors rely on expectations and intentional strategies for organizing. For example, you might look for material that relates to a specific time or place and mentally group that material.

In general, I find it useful to start by thinking of concepts as having components that are related in some way. For example, if we are talking about "learning" we can break it down into organizing the information, entering it into storage, retaining it, and then digging it out when relevant. Each of these components in turn has subcomponents. If I were going to give a lecture on learning, I would devote a part of the lecture to each of the four components. And then within each of these parts I would do the same thing. For example, in discussing the first component, organizing the information, I would talk about bottom-up factors and then turn to top-down factors.

In terms of organizing material for a course, this sort of hierarchical technique makes sense. The course should have "course level learning objectives," each of which corresponds to a "unit" in the course—a set of related lesson plans. Within each of those lesson plans, there should be specific learning objectives, which drive the content. And the content should be organized conceptually, with related material being presented contiguously, "nearby" in time.

A lot is now known about how both bottom-up and top-down factors contribute to chunking. This information can easily be used to help faculty organize material into easily digested chunks.

What is the best way to implement these principles in an efficient way? These principles, although helpful, seem very time consuming.

They should be used at the outset, when a course is first being designed. It's no more work to use the principles to design a class than not to use them – and in fact, it may actually be easier to design classes with the principles from the science of learning, which help to structure the material. The hard part is if you already have a course developed and now want to retrofit it by adding active learning. This is sometimes tricky, in part because you need to delete some material in order to make room for the active learning components. It's best to start from scratch when using the principles to design a course and individual classes.

How has teaching changed over the years?

This question seems to include the assumption that it has changed. In my view, it has changed remarkably little over the years. Lectures are still the predominant form in which information is conveyed in class, and to my knowledge "open discussion" is still the most common format for a small seminar. What has changed is knowledge about how better to make use of these educational settings, how better to use the science of learning to help students learn. Unfortunately, these discoveries (many of which are summarized in my book, "Active Learning Online: Five Principles that Make Online Courses Come Alive") are rarely used systematically—either online or in-person.

How has online teaching changed the teaching game? What have been some challenges and advantages?

I've written a book on this, and won't try to summarize it here. Come the pandemic, many faculty simply tried to teach over Zoom just as they had when they were meeting in person—which was not optimal. Many faculty have since adapted, and started to take advantage of the strengths of the technologies that can be used in online teaching (such as the ability to have small breakout rooms), which in turn now has led to new thinking about how to combine in-person and online teaching as we emerge from the worst of the pandemic.

How do you see Upword applying your principles?

Upword has already been using the principles in creative ways. For example, by asking students to "correct" a machine-generated draft of a summary, the Principle of Deep Processing is drawn upon—students must go back to the material, think about it, and use that information to update the draft. Similarly, in doing this they are led to organize the material into chunks, and to associate what they read to what is already in the summary. I really like the approach Upword is taking, and will bet that students will learn more effectively by going through this process.

For more information about the 5 principles, see blog post: Stephen Kosslyn - 5 principals of effective learning,