A few weeks ago while home from the University of Central Florida where he is studying Electrical Engineering, my son was working to finish up some physics problems before driving back to school. Most of the problems he blew through by looking up the appropriate formula and plugging in information the problem statements provided. One particular problem however, was giving him fits. After struggling with it for about 30 minutes he finally asked old Dad to take a look.
Now it has been way too many years since I took my last physics class. So I turned to Google and quickly located several pages of explanations on how to solve problems similar to his assignment. When he tried following them though, none of them gave the right answer. So much for Dad’s help! Later that evening he texted me to say that he had gotten back to campus safely and had solved the problem.
When home again a few weeks later, I asked him how he figured out the problem. Basically, we were interpreting one of the factors in the problem statement wrong, which led to us putting in the wrong variable. The rest of the story though, was him sharing about the test he had just taken. “Dad,” he exclaimed, “five of the ten questions on the test were related to that same formula and I know I aced them all!” When asked about the other questions, he admitted that he was less than sure about his answers on those and had to look up some of the formulas for those problems.
I think Austin’s experience bears out the conclusions of Nate Kornell, Matthew Hays and Robert Bjork, of U.C.L.A., from their recent study published in the Journal of Experimental Psychology: Learning, Memory and Cognition. Learners remember better and for far longer when they don’t immediately get the correct answer, in other words — when they fail first.
For years, educators, and by extension some training developers, advocated “errorless learning” (first introduced by Herbert Terrace in 1963) where activities were designed to avoid learners getting an incorrect answer. The fear was that making an error (or practicing a skill incorrectly first) would lead to learning the error or inhibit learning the correct information or skill.
The research by Kornell, et al, reflects what we’ve seen in developing training experiences and simulations based on discovery learning. In our Blueline Blueprints‘, where table teams explore information based on a learning visual, the questions where teams have the most debate and spend the most time exploring possible answers create the most memorable learning. Questions where answers are too easily discovered don’t have nearly the same impact or retention.
We see the same in developing engaging and results-producing simulations (whether classroom or technology based). Post-simulation and follow-up testing often shows that learners retain best the material where they first made a wrong choice or incorrect decision. Of course, for the learning to be effective this must be immediately followed by some type of coaching, feedback and rationale for the correct choice.
Just like in my son’s case, when the learner fails first it often invokes a stronger emotional response to the material. Also, the learner spends more time processing the new information and aligning their original thoughts with this correct information. It seems reasonable that this increased energy, focus and attention would result in better learning and retention. And this is just what the research shows.
Of course, to be effective the experience should not just lead to failure willy-nilly. There are some guidelines to keep in mind when designing activities that provide the opportunity to fail.
1.There should be a reasonable path to the failure — The learner should not feel that they were tricked into failing. The given information or context should provide a reasonable frame for the anticipated incorrect choice or answer. However, the correct choice/answer should not depend on too nuanced a rationale. The response we want to generate once the correct choice/answer is discovered is Ah, “I can see that.”
2.Always provide feedback, coaching, or the correct answer — This should come immediately so that incorrect information is not retained. Part of providing this feedback or coaching is to giving a clear and succinct rationale for the correct choice/answer. In some simulation designs, the learner can call on a coach or mentor within the simulation before making a choice she is unsure of. While this may prevent a “failure,” the engagement, emotional energy and time spent processing the material is similar to what would have happened if the coaching had come after an incorrect decision.
3. Provide multiple opportunities for failure, but don’t overwhelm the learner — Look for critical content areas to make the most challenging. If every decision leads to failure the learner gets frustrated and probably will disengage. The most challenging situations or questions should be spaced throughout the learning experience to maintain high levels of engagement.
4.Ensure there are sufficient successes to maintain learner engagement and motivation — This is the corollary to #3 above. While the failures do seem to better ingrain new material, successes provide a necessary boast in energy and motivation. Typically, in designing a 90 minute Blueline Blueprint activity, that may consist of 20-25 discussion cards, I will look to develop 4-5 cards that I believe will generate significant debate about the “right answer.” These cards focus on the most critical aspects of the content. While the other cards are designed to generate discussion and engagement, they are likely to lead to consensus more quickly, and encourage a “yeah, we got that one” response. This response becomes all the more powerful and engaging precisely because the team has struggled and possibly “failed” previously.
Ultimately, all of our training efforts are directed toward helping people be more successful. And a few well-planned failures along the way may be just what it takes to help them ultimately succeed.