Case Study 19.1: The Piano Teacher Who Changed Her Feedback

The Setup

Elena has been teaching piano for twelve years. She's good at it — her studio is full, her students perform well at recitals, parents are happy, students return year after year. She has no obvious reason to change anything.

Then she attends a two-day workshop on learning science and teaching practice, organized by a local music education consortium. The workshop covers retrieval practice, spacing, and feedback. Elena doesn't think the feedback session will apply much to her — she gives feedback constantly. That's most of what teaching is.

But she walks out of that session with something uncomfortable: a new frame for looking at her own teaching, and a suspicion that what she's been calling feedback might not actually meet the definition.

The Observation

Elena spends the week after the workshop observing her own teaching more carefully than she ever has.

She keeps a simple tally: for every piece of feedback she gives during a lesson, she marks whether it is: - Generic positive: "Good," "nice," "that was lovely" - Generic corrective: "Not quite," "again," "try it once more" - Specific positive: Names what specifically worked and why - Specific corrective: Names what specifically went wrong, where, and how to fix it

At the end of the week, she tallies the marks.

Generic positive: 47 Generic corrective: 38 Specific positive: 11 Specific corrective: 19

She stares at the numbers for a while.

Of the 115 pieces of feedback she gave across six teaching days, 85% were generic — either pleasant or discouraging, but containing essentially no information about what actually happened in the student's performance. Fifteen percent were specific. Of the specific feedback, roughly half was positive (what worked) and half was corrective (what didn't).

"I've been making sounds at my students," she later says, "not telling them anything."

What She Was Actually Doing

The workshop had introduced a concept that stuck with her: feedback is only feedback if it contains information that changes future behavior. "Good" changes nothing — the student doesn't know what to do more of. "Not quite" changes nothing — the student doesn't know what to do differently.

Elena recognizes, uncomfortably, that her generic feedback was largely serving social-emotional regulation — making students feel encouraged or making them try again — rather than providing the information they needed to improve. The lesson felt productive because there was constant engagement and feedback-giving. But most of that feedback was nutritionally empty.

The specific feedback she did give tended to cluster around the beginning of pieces (when technical issues were most obvious) and to drop off as lessons progressed (fatigue, time pressure). The most complex musical moments — the places where the most important developmental work happened — were often receiving the least specific feedback.

The Redesign

Elena makes several specific changes to how she teaches:

Change 1: Ask first, tell second. Before providing feedback on any section the student has just played, she asks: "What did you notice about that passage?" She waits for an actual answer. Only after the student has attempted self-assessment does she add her own observations.

This is uncomfortable at first — students are unaccustomed to being asked and often say "I don't know" or give very general responses. She waits through the silence, doesn't rescue them from the question.

Over several weeks, something changes. Students start listening to themselves differently — because they know they're going to be asked. Their self-assessments become more specific, more accurate. A student who once said "it was okay, I guess" starts saying "the left hand was late on the chromatic run in measure 8."

Change 2: Specific, observable, behavioral language. She trains herself to avoid general evaluative terms ("good," "sloppy," "beautiful") and replace them with specific behavioral description.

Not: "That was lovely." Instead: "Your right hand held the phrase all the way through — you didn't break the line before the resolution, which is what gave that section its shape."

Not: "That's not quite right." Instead: "Your left hand is landing on beat 3 ahead of the beat — about a sixteenth note early on the repeated figure in measures 12–14. Listen for where the right hand is when the left hand should land."

The specificity takes more words. It also takes more thought — she has to actually notice what happened, not just notice that something happened. This is harder. It's also more honest: generic feedback is often a cover for not having paid close enough attention.

Change 3: Balance positive and corrective. She notices that her specific feedback had been predominantly corrective — she was better at naming what went wrong than what went right. But students need to know what to replicate as much as what to change.

She makes a rule: before ending any section of a lesson, she names at least one specific thing the student did well that they should consciously try to replicate.

The Results: Student Progress

Elena is a careful observer of her students, and she begins to notice changes over the following months.

Technical corrections hold longer. When she gives specific feedback — "your bow arm is consistently dropping at the tip of the bow on up-bows; keep it level by raising the elbow slightly" — students implement the correction and it persists. When she had given generic feedback — "watch your bow arm" — students might try harder for one repetition, then revert.

Students practice more effectively on their own. This is the change she notices most. Students are arriving for lessons with more specific things to report — "I worked on the passage in the second page and I think I figured out why the left hand keeps being late." They've been having more productive conversations with themselves during practice, because the lesson has modeled what specific attention looks like.

A comparison she can't fully control: She has two students learning the same piece — a Bach invention — in adjacent weeks. She's been teaching the first student (Mei) for three years under her old feedback style, and the second student (Anya) for two months under her new approach.

Their technical starting levels at the beginning of the piece were similar. Four weeks into the piece, Anya has cleaner voice independence and better control of the ornaments. Elena is careful not to over-interpret this — there are many confounds, and two students is not a controlled experiment. But it's consistent with what she expected.

The Deeper Change

The most significant change Elena observes isn't in her students. It's in herself.

Giving specific feedback requires genuinely paying attention to what's happening in the performance — not just whether it sounds right or wrong, but what specifically is producing the sound, at what point in the phrase, with what likely cause. This is harder than generic feedback. It requires the kind of close listening that performance quality requires.

She's become a better listener. And a better listener is, almost definitionally, a better teacher.

"I was giving feedback as a social activity," she reflects. "Now I'm giving it as an informational one. The difference is that I have to actually know what I'm observing."

The Lesson

Elena's story illustrates a pattern that applies far beyond music teaching: most feedback that is labeled "feedback" in educational contexts is actually social-emotional communication — encouragement, correction signals, engagement maintenance — rather than genuine information about what specifically happened and what should change.

The shift to specific, process-oriented, behaviorally linked feedback requires more of the teacher. It requires genuine attention. It requires a language of specific observation. And it requires the discipline to ask "what did you notice?" before telling students what you noticed — building their self-monitoring rather than replacing it.

The payoff is students who can practice independently in ways that actually improve their performance. Which is, ultimately, the whole point.