Teach Better Series: Desirable Difficulties (part 1)

Gamification is academic speak for creating a learning game which attempts to motivate students differently while also making learning enjoyable.  It has tremendous promise and, when performed effectively, can be a (pardon the pun) game changer in the classroom.  But for every epic win in a higher education classroom, there are likely a greater number of fails where gamification did not change the experience in a positive way.  Like MMORPG’s, the education landscape is littered with the bloody and bruised who could not get a classroom game to work.

Why those games did not work is a massive body of evidence in need of a forensic expert.  But it is likely safe to say that some of the problems can be grouped into high-level themes.  Karl Kapp (Gamification of Learning and Instruction, 2012) explains how to avoid many of the missteps in gamification, but also shines a light on typical errors practitioners make.  Some teachers seem to spend too little time truly understanding educational gamification (vs assuming they know what a “game” is) so as to incorporate it effectively into curriculum.  Some instructors attempt to gamify curriculum through points and badges, effectively using the least interesting / motivating part of game mechanics to somehow motivate students.  Some professors create a gamified experience they would enjoy, without considering the generation / student type(s) they are creating the game for.

Similarly, we find a crucial concept to learning as the creation of desirable difficultiesBjork and Bjork gave the grouping of concepts a name and are a few of the current scientific-evangelists regarding this powerful concept.  But many of these concepts have been discussed by educational theorists since Dewey, who argued for learning to be constructive, practical, and reflective.  More and more research from cognitive scientists, education psychologists, and learning specialists have come together in the past two decades to give validation to the power and importance of these concepts.

One of the greatest insights in the last 20 years that has serious potential to improve classroom teaching has been Robert Bjork’s concept of desirable difficulties (Bjork, 1994; McDaniel & Butler, in press), which suggests that introducing certain difficulties into the learning process can greatly improve long-term retention of the learned material. In psychology studies thus far, these difficulties have generally been modifications to commonly used methods that add some sort of additional hurdle during the learning or studying process. (Psychology Today, 2011)

Yet desirable difficulties, much like gamification, are easy to mess up.  While we are starting to see the importance of making learning a struggle to one degree or another, there are easily identified points of possible failure.  For example, practitioners, particularly in higher education, have associated difficult with words like ‘rigor’ or ‘exhaustive.’  That is to say, there is a belief that a curriculum is more valuable if it is larger than it used to be, at a level higher than it once was, etc.  One might even take it so far as to assume less-than-effective means of communication makes something difficult to learn, therefore making that channel or process desirable. 

But desirable difficulty is not about how difficult a concept is, but about how attainment of that concept was administered in reference to how the brain decodes, encodes, stores, and retrieves information.  Desirable difficulties speak as much to instruction as to learning with filters like spaced repetition, retrieval, problem-finding / puzzling over reading, changing learning settings, making learning material less clearly organized, and even using fonts that are slightly harder to read.  All are examples of desirable difficulties.

But it should be noted that some of the desirable difficulties within the education context can be just as uncomfortable for instructors as they are for students.  The example of massed / blocked instruction is an easy observation.  How often does the semester schedule follow a topical pattern of 1 week = 1 topic?  For a single week, all instruction, assessment, and presumed learning occurs around that topic.  It is covered with breadth and depth (or at least as much as 1 week allows), with all focus and attention being on that subject.  Then, the next week, topic 2 is introduced.  Often, topic 1 is not revisited until a mid-term or final exam, with only passing nods given to it as a concept throughout the rest of the class.  It is also assumed by both the instructor and the learner that this topic has been learned, in the whole sense of that term. 

However, Bjork and Bjork proved that making learning too easy and straightforward, as in this type of blocked / massed instruction, can cause a misleading boost in retrieval strength without causing the deeper processing that encourages long-term retention afforded by higher storage strength.  In other words, it may (although not guaranteed) help a student with short-term gains of remembering the concept.  However, long-term retrieval will almost always fail.  So, while it appears that this style of instruction and learning (massed) is optimal to the naked eye, it is actually not useful in any way, with forgetting taking place almost immediately upon exit of the course, if not sooner.  This starts to explain department meeting conversations where upper division professors question lower division instructors about whether or not they “taught” a subject, which students seem to have forgotten by the time they reach the 300 or 400 level class. Of course students forgot the information. 

Known as Interleaving, moving onto a new topic before the first topic is “complete” is not only hard for students, making learning feel more ‘chaotic’ than they are comfortable with, it is also very hard for instructors who struggle to see gains from students in an easily observable way.  So, professors stick with topical, massed instruction as it fits easily into their short-term paradigm instead of doing what is best for student learning holistically.

There is an irony to this concept that may have other cognitive / behavioral science behind it.  Educators sometimes lament the lack of maturity or foresight students have (or don’t have), because they seem unable to look at the importance of any concept for their own future.  The irony is this may not be a maturity issue at all, but a human issue.  Even professors can struggle to choose a real learning paradigm over a known, “fake” learning paradigm.

Researchers have recently discovered just how difficult this concept is for all humans.  We struggle mightily to think about the future in a real, meaningful way.  Humans literally think of their future selves as strangers (at best) or enemies (at worst).  After all, why would current “me” deprive myself of something to save money for future “me?”

That is likely part of what is at work here.  Students struggle to see why long-term retrieval of topics and concepts will matter, just as much as faculty struggle to look past today’s lack of understanding by students to tomorrow’s optimized usage of information instead.  In other words, why would current professor see poor grades on assessments to find better grades on assessments for future professor?  (And shouldn’t this entire concept change how we think of and administer assessments in the first place?)

Not to belabor massed instruction, let us quickly examine a few other desirable difficulties.  Changing the learning settings is another strategy that might prove harder for the practitioner than the student, but is a crucial way to learn.  Again, Medina augments what Bjork, Smith, and Glenberg (1978) found: students need changes to setting, including use of movement, standing, sunshine, and other factors to stay “fresh” as they learn, but also to contextualize and be able to retrieve information without place reinforcement.  Medina goes as far as to finish a section on learning with this passage,

“As I was writing Brain Rules, it hit me [that] if you wanted to design a learning environment that was directly opposed to what the brain is naturally good at doing, you would design something like a classroom.”

Educators are fighting a historical battle with regard to setting.  Classrooms were socially engineered against this exact concept.  The origin of the classroom as we know it was standardized in the late 1800’s by education architects like Cubberley, who said,

“All focus must remain at the front for the teacher [manager], the chalkboard, and the information, with students in rows, quietly reading books, and absorbing information, only speaking when spoken to and only taking breaks when bells ring.  Promoting isolation and fear will create the working class America needs (How Teachers Taught, Cuban, 1984).”

If classroom structure was designed to isolate and create fear, it becomes obvious that the desirable difficulty of changing setting makes sense.  Yet, it begs the question of how often classrooms are still set with desks in rows, a lone speaker at the front, and students quietly reading or directing all attention to one specific place. 

Finally, organization of materials should avoid a “spoon-fed” format, asking students to actively participate in their own planning, organization, and encoding.  This is as true of creating perfect handouts as it is using high definition, high fidelity images during a presentation, even including fonts that require a second look.  Aligning with cognitive load theory and ensuing studies, the following example should be extrapolated to other disciplines:

Say an anatomy professor was teaching students about the heart.  Rather than spending minutes / hours doing a Google Image search to find the perfect, highly labeled, crystal clear picture of a heart, then putting that image in front of students during a lecture, master teachers should take another angle.  Ask students to draw their own heart.  (This is especially workable if using shared whiteboard technology.)  Then, ask students to collaborate.  What seems right or wrong about their peer’s heart?  Ask them to label the most important (say 4-6) things about a heart that a beginner should know.  Then display all the hearts up on a screen.  How many hearts are symmetrical?  Why is that wrong?  How does asymmetry help us understand the heart better, regarding pumping and pulling of blood through the system?  Low-fidelity, less organized methods are a desirable difficulty every professor should leverage.

Do you see the power of disequilibrium, especially when created as a barrier that is not too hard but not too easy?  Do you see the value of making learning “less-easy” so as to produce better results?  For one last example, think about how Dan Meyer approaches math and science.  Once a “whiz kid” of design, this former Jr/Sr High teacher, now a PhD from Stanford, helps K-20 educators create do-first learning.  (That is learning that does not start by telling students what to think, but allowing them to try something.)  By getting students to attempt to answer a question, bet on a condition that does not seem math-based but in fact is, or simply put a guess on the record to a highly complex, variable-rich problem, Meyer is not only leveraging generative learning (a later topic), but also desirable difficulties, leading to motivation and a willingness to try. 

This blog series is designed to help point out, define, and find ways to apply the desirable difficulties, but it is the onus of the instructor to dig deeper.  After all, do no harm is one thing when the harmful behavior is not known to be harmful, but altogether different when it is known but ignored.  Part 1 of Desirable Difficulties was a bit more theoretical, part 2 will “talk turkey” about what this actually starts to look like in the real world (classroom).

Good luck and good learning.