Accuracy first: the one typing-instruction principle worth being strict about

Evidence grade: STRONG by convergence. Direct trials of accuracy-first vs speed-first typing instruction in children are scarce (the field's familiar gap — we grade the benchmarks THIN too), but the principle sits at the intersection of three well-established literatures: automatization encodes what's practiced, error correction costs more than error prevention, and the large typing datasets show accuracy and consistency — not raw key-speed — separating fast typists from slow.

Every typing program makes a governing choice, visible in its stars and leaderboards: reward speed, or gate on accuracy. It looks like a taste decision. It isn't — one of these choices compounds and the other quietly caps a child's ceiling for life.

The case, from three directions

1. Automatization is indiscriminate. Skill acquisition proceeds by compiling whatever is repeatedly practiced (the fossilization argument): practice a wrong-finger reach or a habitual transposition at speed, and that is what automatizes — below awareness, where correction later costs a documented performance dip most learners refuse. Errors aren't neutral events on the way to speed; they're deposits in the same account.

2. The economics of the error. A wrong character isn't one wasted keystroke — it's detection, backspace, retype: three-plus actions and a broken rhythm. The keystroke-level research shows error handling is where untrained typists hemorrhage time; Dhakal et al.'s 136-million-keystroke analysis (2018) found faster typists distinguished less by raw finger speed than by fewer errors and smoother, overlapping keystrokes (rollover). Accuracy isn't speed's rival; at the mechanics level, accuracy is where speed comes from.

3. Speed pressure degrades exactly what needs to consolidate. The speed-accuracy trade-off is one of motor control's oldest laws (Fitts, 1954, and everything since): push tempo and error rates rise. Push tempo during acquisition and you're practicing — hence automatizing — the degraded form. The motor-learning prescription is boring and right: correct execution at comfortable tempo, speed as the product of automatization rather than its forcing function. Musicians have taught this way for centuries; typing is fingering patterns on a different instrument.

Running it without killing the joy

Accuracy-first fails in classrooms when it becomes accuracy-only — drill-sergeant perfectionism that makes typing hateful. The workable version:

Gate, don't grade: set an accuracy floor (we use ~90–92%) that must be met before speed even counts — then let speed stars motivate freely above the floor. The child experiences "get it right, then get fast" as a game structure, not a scolding.
Errors are corrected, not punished: the backspace-to-fix discipline (no advancing past a wrong character) is the teaching — it converts every error into an immediate correction rep (feedback at the keystroke).
Keep drills short and spaced — accuracy degrades with fatigue, and ten fatigued minutes practice the wrong thing (spacing).
Name the deal for children: "slow is smooth, smooth becomes fast" is a sentence a nine-year-old can hold — and it's literally the motor-learning literature.
Watch the fast-and-wrong child specifically — the one gaming speed metrics with 85% accuracy is automatizing errors while being rewarded; they're the child the gating exists for.

What the evidence doesn't say

It doesn't say speed doesn't matter — fluency is the goal; the claim is about ordering, not priority.
It doesn't fix the magic accuracy number — our 90–92% floor is practitioner calibration (high enough to prevent error-automatization, low enough to stay reachable), labelled as such.
It doesn't come from child-specific trials — the convergence argument is strong, but we'd love the direct study, and say so per house rules.

How Wiz Kids applies this

Accuracy is the pass-gate on every typing task (≥90–92% before speed stars are considered), wrong characters must be backspaced before the line advances, drills are short and run on the spaced thread, and the fiction sells the deal in child language. Speed benchmarks exist (the table) — but they're the second star, never the first.

References

Fitts, P. M. (1954). The information capacity of the human motor system in controlling the amplitude of movement. Journal of Experimental Psychology, 47(6) — the speed-accuracy trade-off's foundation.
Dhakal, V., Feit, A. M., Kristensson, P. O., & Oulasvirta, A. (2018). Observations on typing from 136 million keystrokes. CHI 2018.
Logan, G. D., & Crump, M. J. C. (2010). Cognitive illusions of authorship reveal hierarchical error detection in skilled typists. Science, 330(6004).
Anderson, J. R. (1982). Acquisition of cognitive skill. Psychological Review, 89(4) — compilation of practiced procedures, right or wrong.

© Glu IO Pty. Ltd. — Wiz Kids (wiz.kids). Link freely; republication requires permission — see terms. Found an error in our reading of the research? We correct fast: tell any teacher piloting Wiz Kids.