Our Standards

How we verify what your child learns

A homeschool platform is only worth trusting if its content is correct. So we don't just write lessons and hope. Every teaching lesson and every kind of practice question is read cold and re-solved from scratch — checking that the “right” answer really is right — before it reaches your child. Here is exactly how, and what we found.

924
teaching lessons read cold, one by one
~5,000
lesson questions re-solved on the merits
392
practice-question generators audited
~3,800
generated exercises re-derived by hand

The principle

An answer key is only trustworthy if someone re-derived it

Most education software checks that a question is well-formed— four options, no duplicates, an answer that matches the key. That catches typos. It does not catch a key that is simply wrong: a Punnett square mis-shaded, a quadratic factored incorrectly, a spelling question where two of the choices are both real words.

So we set a harder bar. For every question, a reviewer ignores the printed answer, works the problem themselves, and only then compares. If they disagree, the question is held back and fixed before anyone's child sees it. We did this for the entire catalog — math, reading, science, language arts, and our learn-to-read program.

What we check

Three passes, before and after publishing

Every lesson, read cold

All 924 teaching lessons were read end-to-end by a reviewer checking five things: the facts are right, the answer key is right on the merits, no question is ambiguous, the language fits the grade, and there are no typos or broken text. Findings are logged per lesson with a verdict.

Every generator, sampled

Our practice questions are generated fresh each time, so we sampled roughly ten exercises from each of 392 skills across easy-to-hard difficulty and re-solved them by hand — checking the answer, the wrong options, and the hint. We fixed 25 generator bugs this way, including a few that could have marked a correct child wrong.

The progress engine, simulated

We run scripted “students” — a struggler, an ace, a random guesser, a kid who bounces back — through the system on every release to confirm it never advances a child who is guessing, never traps a child who is struggling, and schedules review at sane intervals. These simulations are permanent automated tests.

What we found

We'll be straight with you: we found things, and fixed them

No honest review comes back perfectly clean, and ours didn't. The answer keys held up extremely well — across thousands of re-solved questions, the displayed answer was right on the merits in the overwhelming majority. The defects we did find were mostly in the supporting cast: a distractor that was accidentally also correct, an explanation that referenced the wrong number, a rendering glitch that showed a raw symbol instead of a fraction.

Each one was fixed at the source and re-checked before this page went up. We think a company that tells you what it found is more trustworthy than one that claims perfection — which is why this page exists at all.

If you spot something

There's a “report a problem” button on every question

Verification is never “done.” Every question in the app has a one-tap way to flag it, and we track which questions get answered unusually often or flagged by parents. Your eyes make the catalog better — and you'll always know exactly how we check our work, because it's written down right here.

See the content for yourself

Try a real lesson, then add your child and sit with them for the first one. Free forever for one child. No credit card required.