Most learning apps let kids succeed without learning.
You've suspected this. Now you can prove it.
One question exposes any broken app
"Can my kid succeed at this without actually learning?"
If yes, the app is broken — no matter what the dashboard says.
Is this you?
Who this is for
Parents who suspect something's wrong
You bought the app, watched your kid use it, saw no improvement. Now you'll understand why — and what to look for instead.
Builders who want to make apps that work
These are the principles. Follow them and your app will actually teach. Ignore them and you're just building another engagement trap.
Teachers who need language for what they see
"My students game these apps" isn't convincing to admin. "The app violates invariant #1" — with evidence — is.
Why apps fail
Kids are cognitive misers
This isn't a criticism — it's how human brains work. We instinctively find the path of least effort. If there's a shortcut, kids will find it:
Guess from pictures instead of reading
The app asks "Which word means 'dog'?" with a picture of a dog next to one option. The child taps the picture. "Correct!" The child never read the word.
Tap randomly until they get lucky
Four choices, 25% chance. Wrong answers cost nothing. Faster to guess than think. The app records "attempt" as "practice."
Use hints as answer-reveals
Stuck? Tap hint. The hint shows the answer. Tap answer. "Great job!" The child learned that the hint button gives free answers.
Pattern-match without processing
After seeing 5+3=8 three times, they recognize the pattern. Ask them "What's five plus three?" verbally and they can't answer.
The app shows "mastery." The child learned nothing.
This is the gap between performance (doing it in the app) and learning (being able to do it anywhere).
The framework
11 invariants that predict if an app will teach
Non-negotiable principles from 40 years of learning science. Each one closes a loophole. Together, they're airtight.
The target skill must be the only path to success
If a child can succeed by guessing, tapping pictures, or using hints — the app isn't teaching. The cognitive work that causes learning must be required, not optional.
Active production is required
Recognition is not recall. Tapping the right answer from a list is not the same as producing it from memory. Learning happens when the brain generates, not when it recognizes.
Mastery cannot be bypassed
If kids can skip ahead, watch videos instead of practicing, or unlock content without demonstrating skill — the app has no quality control on what enters their head.
The foundation
Built on 40 years of learning science
These aren't opinions. They're principles derived from decades of cognitive science research — the same research that most ed-tech ignores because it's harder to build apps that actually work.
5-10%
Typical efficacy rate across ed-tech. Most apps don't move the needle.
11
Invariants that separate apps that teach from apps that entertain.
Binary
Pass all 11 or fail. No partial credit. No "it depends."
Once you understand these principles, you can't unsee them. You'll watch your kid use an app and immediately spot where the learning leaks out.
Start with the methodology
The 11 invariants are the core. Everything else builds on them. Read them once and you'll never look at ed-tech the same way.