New findings from ongoing research by a team of American researchers at Johns Hopkins University and The Kennedy Krieger Institute shed some more light on this topic with implications for athletes and coaches from all sports.

The results of their latest study, published in the journal Current Biology, demonstrate that, when it comes to making adjustments to established movement patterns, trial and error, although more time consuming, leads to longer-lasting changes than the rapid but short-lived changes derived from external feedback.

Lead author Dr. Ryan T Roemmich talked to World Rowing about these findings and what it could mean for how coaches and athletes think about making changes that will stick.

Adaptive learning

“We studied a form of learning called locomotor adaptation, or adaptive learning,” says Roemmich, who holds a PhD in Biobehavioural Science. “Adaptive learning mechanisms are triggered whenever we receive sensory feedback that doesn’t align with the feedback we expect from our movement.”

To illustrate what he means by adaptive learning, Roemmich uses the example of swinging a baseball bat with a weighted ‘donut’ on the end (a rowing example might be putting some extra load on the oar). In either case, the athlete can immediately feel that something is different and has to change the way he or she executes the swing of the bat (or drive phase of the stroke) to move it at the desired velocity.

“This recalibration is adaptive learning taking place,” Roemmich notes. “You are learning to use a familiar tool in unusual conditions.”

In rowing, something as simple as variable water conditions can lead to many such small-scale situations even within a single session.

“These are situations where you learn from the mistakes that you feel. For example, ‘the bat or oar feels heavier than I expected, the water is harder to push than I expected.’ This type of learning we find to be uninfluenced by external feedback (as may be given by a coach),” says Roemmich.

However, while immediate and ongoing feedback during this sort of learning does help the athlete make changes faster, once the feedback stops, so does the improved technique. This is unless the athlete has simultaneously been ‘feeling’ through trial and error how things have changed.

Learning to change

Although their findings are highly applicable to adaptive learning situations across sports, Roemmich’s team looked specifically at movement on a split-belt treadmill. Each belt could be sped up or slowed down independently, creating unfamiliar situations where they monitored the ways people learned to change their familiar patterns of movement.

“Over time, [the participants] adaptively learn to eliminate the limp and walk symmetrically despite the asymmetry in belt speeds,” Roemmich explains. “The adaptive learning occurs because the outcome of the movement differs from what was expected – people expect the belts to move at the same speeds and walk accordingly, but adapt their waking patterns to accommodate the new environment.”

The researchers then introduced a source of external visual feedback to assess if learning improved over simply having participants feel out the changes on their own. “Walking symmetrically is more stable and costs less energy than limping,” says Roemmich. “So we thought that providing visual feedback may help people to acquire this more desirable walking pattern faster.”

While the learners did make changes more quickly with visual feedback, the changes disappeared once the feedback was removed. “This indicated that people learn to eliminate their limp by feeling the limp and correcting it themselves, while the visual feedback simply provided a sort of ‘quick fix’ that was not maintained once we took the feedback away,” says Roemmich.  

The end of coaching?

Fortunately for coaches feedback still has an essential place in athlete skill development. “Our results do not suggest that external feedback or coaching is bad for skill learning,” Roemmich says. “Instead, they suggest that coaching – while it can drive faster changes in movement – is not particularly effective for accelerating adaptive learning.”

The best method, according to Roemmich, may be a blend of both. “Many studies have suggested that providing intermittent feedback drives the longest lasting improvements in skill,” he says. “Others have also looked at the practice structure itself and found that variable practice – or practicing different movements in short, alternating intervals – can also lead to longer lasting improvements.”

“The beauty of adaptive learning,” says Roemmich, “is that it lasts even after the stimulus driving the learning has gone.”

Looking back to our rowing example of added load, while the boat feels heavy and the movement different at first, once the rower’s movements adapt, the boat feels incredibly light when the load is returned to normal (eg half boat rowing followed by full boat rowing).

“Adaptive learning inherently has lasting effects because a well-established movement pattern has been subconsciously recalibrated,” Roemmich says. “Because it’s subconscious, that recalibration lasts even upon return to a normal environment.”

It seems that there is more truth than ever to the old proverb: tell me and I forget; teach me and I may remember; involve me and I learn.