The heart of our pedagogical approach is the theory of “failure-driven learning,” which says, in a nutshell, that learning happens only when the world does something different than what you expected.

In the upbeat, positive-thinking world of instructional design, the phrase “failure-driven learning” strikes some people the wrong way. It sounds, let’s face it, a little negative. That was presumably what motivated the person who, at the end of a presentation I gave on the subject, asked, “Why not success-driven learning?”

I think that question is worth answering, not just because it’s kind of clever, but because I think it neatly captures the most common objection to failure-driven learning as a theory.

To respond to this, I first want to point out something that might be obvious, but that can be confusing. The “failure” in failure-driven learning refers to the failure of an expectation. It’s not the failure to execute a plan or achieve a goal. Those kinds of failures often happen as the result of an expectation failure, and the desire to avoid those kinds of failure provides motivation for learning. However, it is expectation failures per se that signal leaning opportunities.

So the question becomes: why does an expectation have to fail before you can learn something new? The answer is that until an expectation fails, there is no way for your brain to identify what needs to be learned.

A successful expectation, desirable as it may be, brings no new information. Consider a trivial example. Suppose you go into a McDonalds expecting to be able to order a Big Mac, only to find that… you are able to order a Big Mac. What is there to learn from a “success” like this? Nothing.

In contrast, suppose you go to a McDonalds one morning expecting to be able to order a Big Mac, only to find that you cannot. Now your brain has something to work with. To improve on your original expectation that you could always get a Big Mac at a McDonalds, you need to try to figure out why it wasn’t so this time. To do that you might notice that everyone else seems to be ordering breakfast food, and infer that the menu is different at breakfast,  Or, you might ask and be told this by an employee. Either way, you will now be able to adjust your expectations so that you know in the future you cannot count on getting a Big Mac at nine in the morning.

Or suppose you go to a McDonalds in Gurgaon, India at lunch time expecting to be able to order a Big Mac, only to find that you cannot. Once again, your mind will seek to explain this expectation failure. You will probably figure out that the McDonalds menu in India is not the same as it is in the US. If so, then you will expect that other dishes will be different as well. You might be able to reason that one big difference is likely to be the absence of beef, seeing as how the majority of the population of India does not eat beef. You might also be able to reason that since many countries have very different cuisines than the  US,  McDonalds’ menus might be different in other countries as well (which is correct). You see in these instances how your ability to reason about why an expectation failed allows you, potentially, to generalize to new refinements of your original expectation, and more general changes to other expectations as well, like predicting that you will not find a Quarter Pounder in an Indian McDonalds, and that you may not find a McRib sandwich at one in Malaysia (actually, there is a McRib there, but, oddly, it’s made of chicken instead of pork).

A failed expectation triggers learning mechanisms in your brain to analyze why this particular expectation failed in this particular context, and the explanation you arrive at forms the basis for new and improved expectations for the future. Which is what learning is all about.

A successful expectation does not stimulate your brain to activity in the same way. This observation is born out if you think for a moment about what happens in highly familiar environments in which all of your expectations are generally met: making breakfast, feeding the dog, driving to work, entering an expense report, and so on. When you are in one of these hyper-familiar environments, what happens? If you are like most people, your mind wanders. You do what you need to do on “autopilot,” while thinking about other things entirely. Psychologists call this effect habituation. When you habituate to something, your mind is in the opposite of a “ready to learn” state while you are doing it.

So the theory of failure-driven learning rests on this very basic asymmetry: Successful expectations don’t yield new insight, don’t frame a learning opportunity, and don’t invoke the brain’s learning mechanisms. Failed expectations do.

And that’s why it’s “failure driven learning” and not “success driven learning.”

Click here to check out part 2 in the series, Failure-Driven Learning vs. Coloring Inside the Lines.