1. RNP
  2. Motor Learning
  3. Performance, retention, transfer
  4. Performance and motor learning: what a successful session doesn't prove

Motor LearningSeries 7/9

SeriesPerformance, retention, transfer7 of 9

Performance and motor learning: what a successful session doesn't prove

A successful session doesn't prove motor learning. What the reviews say about in-session performance and retention, and how to validate a movement cold.

LabO RNP·10 min read·reviewed on

Day 5 of our book, and what the reviews say about in-session performance and retention.

By the LabO RNP team

Motor learning: what's left two days later, when the coach is gone

Hello, movement professional,

A weight room, just past six in the evening, plates clanging on the next rack. A javelin thrower is working in a lunge, block leg forward, a light bar held overhead with arms locked out. The position prepares the shoulder to come through high at the moment the javelin leaves the hand. Under his elbow, the strength and conditioning coach's hand rests, palm open. It isn't holding anything up; it's guiding. On every rep, it lifts the elbow by a centimeter while the coach keeps repeating “tall, push the ceiling.”

The bar settles a little better from one rep to the next. By the sixth, it doesn't move at all. The throws coach walking past behind them drops a “clean,” and the S&C coach takes his hand away to watch the thrower do it alone. The first rep holds, the second more or less. On the third, the elbow bends, the bar drifts forward and comes down as if the session had never happened.

People will say the guy was fried, that it was good work and it'll come back on Monday. You'll find that third rep in every gym, and every time fatigue gets the blame. It's the typical case that opens day 5 of our book, and it raises a question almost nobody asks at the end of a session: of what you just saw, what will hold without you? Motor learning research answered that long ago: what someone produces at the end of a session is a poor predictor of what they'll reproduce later, without a reminder (Kantak and Winstein, 2012; Soderstrom and Bjork, 2015).

Why we judge a session by what we see at the end

You run your session, the person you're working with improves before your eyes, you sign off on it, and everyone goes home happy. Underneath all that is a rule we all apply without ever having written it down: if it's better at the end than at the start, it's been learned.

It holds up for good reasons. A successful end of session can be shown to a colleague, a parent or a coach without any explanation. It reassures you, first of all. And nobody argues with a session where the athlete finished better than they started.

Except that performance is what you observe that day, in that room, with the help you were giving at that moment. It tells you nothing about the following Monday.

The RNP training space, free for life

Get access to RNP training, no strings attached.

Welcome! Introduce yourself

SecureNo cardNo password

By continuing, you agree to our Terms of Use and our Privacy Policy.

What drives in-session performance up without guaranteeing learning

Improvement during a session can come from several sources. More frequent feedback from you pushes it up. So does a more precise cue, or a task you simplified without realizing it. The person's attention goes up because you're watching them. A compensation gets a little better tuned over the set. All of these sources raise what you see at the end of the session, and none of them guarantees that anything will be left by the next one.

The end of a session is a snapshot, and you're in the frame. Learning shows up in a film, several sessions strung together, in which you no longer appear.

The trap works both ways. It gets you to validate whatever shines, and it gets you to drop too early the exercise that makes the session look less pretty, even though it may have been the one leaving something behind. We've all fallen for it, us included, dozens of times.

Learning versus performance: what the reviews say

Studies separate two measures. Performance is what the person produces during practice. Learning is measured later: retention, what remains when you come back with no practice in between, and transfer, what carries over to a task or context that wasn't trained, all the way to far transfer when that context is very distant.

In 2012, Kantak and Winstein reviewed the studies comparing the end of practice with a test repeated later [1]. That's the snapshot compared with the film. Their conclusion: retention and delayed transfer tell you more about learning than end-of-practice performance does. The gap is clearest when demanding practice degrades the session while improving what remains. Of the whole book, this is the point where the studies converge most strongly.

Soderstrom and Bjork reached the same conclusion in 2015, for verbal as well as motor learning [2]. What you observe during training is a poor indicator of what has been learned, and some ways of structuring practice even have opposite effects on the session and on what remains. They add that learners often mistake their performance on the day for a reliable guide to what they've learned. So the thrower who finished with the bar locked in place is just as convinced as you are that he learned something.

The hand under the elbow and feedback on every rep

For the number or comment you give after a rep, research first pointed in the same direction. In 1990, Winstein and Schmidt showed that with feedback on only some of the reps, the movement held up better on the delayed test than with feedback after every rep [3]. That result served as a rule for a long time. A 2022 meta-analysis found no significant effect of reduced frequency, either during practice or on the delayed test, and judged the included studies to be severely underpowered [4]. So we keep the principle as a practical lead, not as a law.

The book goes further, and says so: in practice, we extend this result on feedback frequency to the hand under the elbow and to the coach's mere presence. It's an openly acknowledged extrapolation that nobody has tested in that form.

Why the movement was resting on the coach

Go back to the thrower's bar. He's aiming for something, he acts, and feedback information comes back to him: the floor under his foot, the bar shaking or not. That's the sensorimotor loop. In the ecological reading we follow in the book, learning a movement means seeking out the right information, then scaling the movement to what that information says. That second adjustment is called calibration. The error is the distance between the bar the thrower was aiming for and the one he actually feels above him. From one rep to the next, that distance is what drives the loop, provided he perceives it himself.

Your presence enters the loop as one more source of information, and the easiest one to read. Your voice arrives before the athlete has had time to feel the bar. Your hand under the elbow erases the gap before it even exists. The loop does adjust, but it adjusts to you.

So the learner learns to read the moment you speak, your word, the pressure of your hand, even the way you position yourself. None of that exists outside your gym. The day you're not there, the movement has nothing left to hold on to, and it doesn't carry over to another gym or in front of another coach either.

It starts early: a children's movement educator told one of us about a five-year-old boy who got the ball into the box every time the educator was crouched beside him repeating the instruction. The next day, with another adult or another ball, the boy waited to be told. The ball in the box was resting on the educator, exactly as the bar was resting on the S&C coach.

A dip during the session isn't always a failure

Performance rises fast when you talk on every rep, because you're handing over the solution and the learner has nothing left to search for. In the other direction, we add a field observation with no study to back it: a dip during the session may be the moment the person lets go of a solution that worked in order to look for one that holds.

That doesn't make every dip a good sign. To tell the two apart, there's only one judge: what comes back cold, or in front of someone else.

Validating a movement cold: the test and the number nobody writes down

The principle itself doesn't change: retest cold, without a reminder, at the earliest at the next session. No study sets a standard delay; two days is a good rule of thumb.

When you do speak, tell the person where the bar ended up rather than what their elbow should do. That way there's still something left for them to adjust, and they're the one adjusting it.

To validate, write a number next to what the person succeeds at in front of you: how many pieces of feedback it took for them to get there. Almost nobody writes it down. Yet that's the number you bring down, session after session, by watching what still holds when you stay quiet.

What is established, what is assumed

Established: the distinction between performance and learning. End-of-practice performance poorly predicts retention and transfer [1, 2], and learners themselves are often wrong about what they've learned [2].

Still fragile: the effect of reduced feedback frequency. Shown in 1990, it was not replicated by the 2022 meta-analysis, which judged the included studies to be severely underpowered [3, 4].

Field observation, presented as such: extending this result to the hand under the elbow and to the coach's presence, the in-session dip as a possible reorganization, and the two-day delay before retesting.

The fact that a movement falls apart when the coach steps away doesn't prove the hand is the cause: real fatigue, context and task difficulty also play a part. The cold test shows what holds for a given person, and the cause has to be found case by case.

When a movement only holds under your hand

The thrower's movement works as long as the S&C coach's hand and voice are there. In-session performance can also be a well-tuned compensation, and what we've just covered isn't enough to spot it. Look at what feeds into the loop, the way an RNP reading does: the inner ear, the eyes, the skin and the joints of the foot reporting the load, in other words proprioception and its neighbors. That's the route the feedback information takes.

If one of these inputs gives poor information, the learner adjusts to whatever is still coming through clearly, and your hand under the elbow stands in for what their body isn't reporting. The session shines, and the loop never received what it was missing. The distinction between calibration and compensation is there to sort this out. When a movement only holds under your hand, ask yourself what the person isn't perceiving before asking what they don't know how to do. This reading is covered in detail in our article on the sensorimotor loop.

RNP is not a medical practice: it complements health care, it doesn't replace it.

Take your hand away and count what holds: the rest was you.

This lesson is day 5 of 28 days to learn how to make people learn, a LabO RNP book in 28 short lessons, one a day.

👉 I want to know what's left of my sessions: the free course “How we learn to move”, whose chapter “Three things you believe about learning” covers the belief “What you see in the session.”

Frequently asked questions about learning and performance

What is the difference between performance and motor learning?

Performance is what the person produces during practice, with whatever help is available at that moment. Learning is what remains when they come back later without any practice in between (motor retention), and what carries over to another task or another context (transfer of learning).

Why doesn't a successful session prove that learning happened?

Because several factors boost same-day performance without guaranteeing that anything will stick: frequent feedback, more precise instructions, a simplified task, heightened attention, a better-tuned compensation. The reviews by Kantak and Winstein (2012) and by Soderstrom and Bjork (2015) show that end-of-practice performance is a poor predictor of what is retained.

How do you test what has been learned?

Retest cold, with no reminder, at the next session at the earliest, using the same criterion as the last successful attempt. No standard sets the delay; two days gives a rough order of magnitude. Recording how much feedback was needed completes the picture.

Should you stop correcting during the session?

No. The book suggests giving feedback on the outcome (where the bar ended up) rather than on the movement itself, and gradually reducing the amount of feedback over the sessions. The book keeps only one case where you step in immediately: danger.


References

[1] Kantak S. S., Winstein C. J. (2012). Learning-performance distinction and memory processes for motor skills: a focused review and perspective. Behavioural Brain Research, 228(1), 219-231.

[2] Soderstrom N. C., Bjork R. A. (2015). Learning versus performance: an integrative review. Perspectives on Psychological Science, 10(2), 176-199.

[3] Winstein C. J., Schmidt R. A. (1990). Reduced frequency of knowledge of results enhances motor skill learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16(4), 677-691.

[4] McKay B., Hussien J., Vinh M.-A., Mir-Orefice A., Brooks H., Ste-Marie D. M. (2022). Meta-analysis of the reduced relative feedback frequency effect on motor learning and performance. Psychology of Sport and Exercise, 61, 102165.

€0Your free RNP PathsStart