Let’s try an experiment.  I’d like you to think about two things:  a book that that you read in the past month and your most recent trip to the grocery store.  Got it?  OK, now, answer these questions:

• For which one can you remember the most specific details?
• Which one was more engaging and interesting?
• Which one are you more likely to remember 5 years from now?
• Which one is more likely to influence the way you think or behave?

My guess is that the book won hands down.

Yet, according to many of the “best practices” that people espouse about learning, the grocery store should have won.

The grocery store is active; you were “doing” 100% of the time.  It was non-linear and provided you with 100% control over the experience (except for the checkout).  The grocery store allowed you to “interact” with your environment.

So how is it that a passive, linear, static, “presentation” that provided no user control or any “activities” or interactions is more effective at influencing what you remember, think and do?

That answer is simple.  You learn from thinking, not doing.  Learning occurs when our current way of doing or thinking is no longer sufficient to achieve a result and we are provided the opportunity to recalibrate our mental models.  Roger Schank called this moment an “expectation failure”.  Schank is also a key proponent of Learning by Doing.  However, his “doing” is markedly different from most of “doing” that I see built into corporate training.  His works because it is designed well and causes you to think.  Most “doing” in training simply has people go through the motions of performing a task.

Learning by Doing came about as a reaction to “Learning by Lecture”.  Traditional lectures don’t work.  That’s because most traditional lectures don’t cause you to think.  They just provide information.  Amassing and storing facts is not learning.  Learning is figuring out how those facts fit together and how they influence other things.

While having someone “do” something increases the chance of learning, it’s no guarantee.  When you give someone a task, you force them to apply their existing understanding of the world.  This creates the opportunity to have their understanding or expectations fail.  However, unless that situation is well designed, they may not actually have to think very much.  After all, if we really learn from doing, we should all be geniuses as we spend more than 80% of our time at work “doing” things.  The problem is that most of what we do (whether in work or in training) can be done on auto-pilot.

For example, having someone fill out an employment form doesn’t teach them how to fill out an employment form.  From their past experience, they probably already know how to perform each task (e.g., interpret the fields on the form and fill them in). They are just applying what they already know to a new form.   More difficult tasks like interviewing or other decision-making often can succeed simply by applying prior experience.

If the activity that someone is doing simply has them apply their existing knowledge and that knowledge is sufficient to complete the task, they aren’t learning.  That’s true even if they are applying that knowledge in a new context.  They only learn when their old understanding doesn’t work in the new environment.

In fact, if your training makes too much sense to your audience, they probably already knew what you were teaching.  In other words, if people complete your activities correctly on the first try, they probably didn’t learn anything.  More importantly, if they only get one pass at the situation and don’t have a chance to reframe and reapplying their thinking, they didn’t learn either.

Good training should initially confuse people (not fool them, confuse them).  It then needs to help them understand where that confusion came from, allow them to re-adjust their mental models and try again to see if their new models work.  This is where Schank’s approach differed from many others.  His design processes takes the time to find out the common mistakes people make or misperceptions they have and builds the “doing” part specifically around those situations.

The key is to create an environment that causes someone to think, predict, and form relationships in their mind.  That certainly can be done with a well-designed activity.  However, it can also be done with a great example or a well told story.  In fact, a good story told in an effective manner will drive greater learning than an activity that has someone go through the motions of applying their existing knowledge.  I bet that your people would learn a lot more from an 18 minute TED talk on sexual harassment than they do from the “activity-based” Sexual Harassment compliance training that you put them through.  That’s because good TED talks engage your mind.  They cause you to think and challenge your assumptions.  And, they do it without a lot of slides and with few to no activities.

Good learning must be purposely and rigorously designed.  It must challenge a person’s current perceptions and understanding of how things work.  Stop counting the number of minutes spent in activities.  Most activities in training, like most activities in real life don’t require much thought because they can be accomplished simply by applying past experience.  When that’s the case, there is no learning happening.

Assess the quality of your training based on how much of the time people spend thinking (not listening), how many decisions/predictions (implicit or explicit) you are having them make, and how often their decisions are wrong (and be sure you provide them with feedback, support, and the opportunity to reframe and adjust their thinking and try again).  That will tell you how effective the learning experience will be.