Hardest Problems Can Make the Best Teachers

For those challenges that are difficult to describe, tackling the hardest example first may be the best way to master a concept.
Publish date:
Social count:
For those challenges that are difficult to describe, tackling the hardest example first may be the best way to master a concept.

To learn a craft, the rule of thumb is start simple and build up to the more complex components. But surprising new research suggests that, for certain types of skills, the opposite approach works better.

An explanation of why begins with a question: Have you ever wondered why your teenager can perform brilliantly at video games but has a C average at school? Research reveals that mastering the two activities involves completely different learning systems in the brain.

Comprehending information imparted in a lecture or textbook requires cognitive reasoning skills, which according to some estimates don't fully mature until our mid-20s. In contrast, developing the ability to score points on a video game uses a more primitive form of pattern recognition — a type of learning that evolved much earlier in our development as a species and matures more quickly in an individual brain.

It's highly unlikely your teen could explain to you why he is able to exterminate so many virtual opponents; the intuitive skill he is using is not something he (nor anyone else) can verbalize. But it's part of our mental makeup and clearly of value, even in a society where logical reasoning rules.

And according to veteran research psychologist F. Gregory Ashby, it literally can save lives.

In collaboration with Brian Spiering, a colleague at the University of California, Santa Barbara, Ashby recently published a study comparing "information integration" (that's the primitive method of acquiring understanding) with "rule-based category learning" (the use of logic and reason). It describes an experiment in which a group of students sat down at computer screens and viewed images of small round discs with black and white stripes.

As the discs rapidly flashed onto their screens, the students were instructed to assign them to one of two categories, using a distinction they had to discern as the experiment proceeded. Sometimes the categories were ones they could describe verbally and figure out logically — say, one set of discs had thin bars and the other had thick bars. In such cases, the participants used conscious reasoning to make their choices (a strategy made evident by the pattern of their picks).

At other times, the discs were divided into categories that were not easily described in words and, thus, not easily discerned through logical reasoning. In these cases, the participants gradually gave up trying to consciously figure out the pattern and let their primitive learning system do the work.

"If you asked them afterwards how they did it, they couldn't tell you," Ashby said. "We got responses like, 'I just went with my gut reaction.' A bunch of people told us, 'I just started humming or singing to myself, and I started getting the correct answer.' "As they were distracting their conscious minds, their unconscious was solving the puzzle.

The series of discs were presented to the students in one of three conditions. Some got the easiest problems first, after which they grew in difficulty; others had them presented in the reverse order, while a third group had them presented at random.

"When there was a rule that could be described verbally, there was no difference (in success rate) in any of the conditions," Ashby said. "That was a little surprising, since there is literature suggesting if you start with the easiest first, there might be an advantage. We didn't find that.

"When you couldn't describe the rule, there was a huge advantage for starting with the most difficult items," he added.

Initially, the participants did very poorly, but within just a few minutes — when we transitioned to the medium level of difficulty — they had a clear advantage, and they maintained it throughout the whole session. We were surprised by how large the effect was."

The results confirmed Ashby's hypothesis. "If you give people the easy problems first, they get rewarded for using simple strategies," he said. "When they get the more difficult ones, they're reluctant to give up what has been working. That actually hurts them."

In contrast, the people who got the toughest problems early on quickly "gave up trying to figure it out using reason," he said. That allowed their primitive, intuitive learning system to kick in.

Ashby is quick to concede these results have limited implications for the classroom — which, after all, is a forum for logic-based learning. But he believes it is useful information for people in mentorship roles — those who are teaching the fine points of a specific skill that requires instinctive knowledge, such as becoming a master chef.

In many professions, "there is a nonverbalizable component that is very important," Ashby noted. "An example we often use is a radiologist reading X-rays. Suppose you are looking at a mammogram and trying to find a tumor. You can go to medical school and hear lectures on the subject.

"But to become an expert, you have to do a residency and work beside a true expert radiologist. That person can look at a mammogram and be very confident of whether there is a tumor there or not, but he or she would not be able to write down a set of instructions that would allow you or I to do the same."

Such training needs to be one-on-one, with the novice making tentative diagnoses and getting immediate feedback (not unlike the feedback the teen gets from his video game, which instantly tells him when he has missed his target). This is the way the primitive learning system operates; it's how a craftsman develops a feel for his craft.

So, if future research confirms Ashby's conclusions and finds them to be widely applicable, it could change the way medical residents are trained — along with anyone else who is learning firsthand a skill one can't fully describe in words. Throwing apprentices into the deep end of the pool — say, by having them diagnose a tricky, borderline mammogram right off the bat —may in fact be the best way of imparting this invaluable type of knowledge.

Sign up for our free e-newsletter.

Are you on Facebook? Become our fan.

Add our news to your site.