A student has been practicing piano for three months. They have not missed a single day, and their teacher praises their consistency. Then one day a thought creeps in — "Am I actually getting better?"
If that question is hard to answer, a measurement system is missing.
People Are Not Good Judges of Their Own Skill
Psychologists Justin Kruger and David Dunning (1999) documented a pattern that is especially pronounced in unfamiliar domains. Beginners tend to overestimate their abilities, while intermediate learners often underestimate their growth. Both errors share the same cause — there is no internal benchmark for judging what you do not know, or what has improved.
The same holds in music. Whether today's practice was better than yesterday's, or which areas have changed since last month, cannot be determined by feel alone. Fatigue and fluctuating confidence cloud the judgment.
Three Time Scales, Three Questions 🔍
Splitting diagnosis into three time units makes it possible to answer different questions.
Daily diagnosis: Which items generated the most errors today?
- Reaction speed, accuracy, specific note regions — the daily scale catches concrete error patterns.
- Five minutes is enough. A note like "got stuck three times on the notes above the fourth line in bass clef" is sufficient.
Weekly diagnosis: Did the trouble areas from last week improve this week?
- This is the scale for confirming growth. Trends that are invisible day to day appear clearly across seven data points.
- Use the same test each week — for example, how many correct in Level 2, Sub-level 1 within three minutes — and run it the same way every time.
Monthly diagnosis: Compared to one month ago, which areas grew most and which least?
- This is the scale for strategic adjustment. A conclusion like "notes near middle C are much faster, but ledger-line notes are still slow" is only visible across a month of data.
The Conditions for Measurement to Mean Something
Two things are required for diagnosis to be meaningful.
First, consistency of method. Testing with different pieces at different difficulty levels each week makes comparison impossible. The same level, the same type of problem, the same time window — only then can last week and this week be compared.
Second, the habit of writing numbers. "Today felt good" is a memory that fades quickly. "Today: Level 3 accuracy 78%, last week 72%" does not fade.
How Digital Tools Changed This Process
In the past, this kind of record had to be kept by hand in a practice journal. Now, the practice itself generates the data. Note recognition training apps automatically save accuracy, reaction time, and weak-note distribution after every session.
Noteflex's dashboard visualizes this data at daily, weekly, and monthly scales — designed to answer the three questions above without the user doing any manual recording.
If that three-month student had been using this system, they would see something on their screen today: "Bass clef accuracy up 14% compared to last month." That is the answer.