Why we don't grade the projects

A kid finishes a project in our room and nobody hands them a grade. No rubric score, no A-minus, no percentage. Parents sometimes notice this and ask about it directly — not as a complaint, usually, but out of habit. Every other structured activity their kid does comes back with a number attached. Ours doesn't, on purpose.

What a grade actually measures

A grade is a compression of a lot of information into one symbol, optimized for comparing kids to each other or to a standard, fast. That's genuinely useful in a school setting, where a teacher has 30 kids and a district needs a number to track over years. It's not useless. But it measures conformity to a rubric someone else wrote, and the thing we actually care about — whether a kid built something real and can explain the decisions behind it — doesn't compress into a number without losing the part that matters.

We wrote about this from the parent's side already, in The output is the rubric: the finished, working thing is the evidence, not a score describing the thing. This post is about why we hold that line even when it would be easier not to.

The actual alternative: does it work, and can you explain it

Instead of a grade, a project here gets two questions. Does it work — does the robot actually avoid the obstacle, does the story actually finish, does the trailer actually cut together? And can the kid explain why they made the choices they made — why this sensor, why this plot structure, why this shot order? Those two questions produce a much better picture of what a kid learned than a rubric score does, and they produce it in the kid's own words instead of an adult's compressed judgment of the kid's words.

This is also just harder for mentors than filling in a rubric would be. A rubric is fast and defensible. "Ask the kid to explain their choices and actually listen to the answer" takes longer and produces something you can't average across a spreadsheet. We think that's the right trade.

The cost we're accepting

Not grading anything means we can't hand a parent a tidy report card, and some parents want exactly that — a number to compare against last month's number, evidence the money is producing measurable progress. We don't have that, and we're not going to manufacture a fake one to make the absence less uncomfortable. What we can offer instead is the actual project, which a parent can look at and judge for themselves, and a mentor's honest read on where a specific kid is stuck or strong — a conversation, not a score.

That's a worse fit for anyone who specifically wants standardized measurement. We'd rather say that plainly than pretend our approach is a superior version of grading. It isn't a version of grading at all.

Why this matters for the kind of motivation we're going for

A kid who's optimizing for a grade is optimizing for the grader's approval, not for the thing itself. We see that shift happen in real time sometimes — a kid asks "is this good?" meaning "will you approve of this?" instead of asking whether it actually does what they wanted it to do. Removing the grade removes the target that question was aimed at. What's left is a plainer question: did I build the thing I set out to build, and if not, what's still broken. That's a more useful question for a ten-year-old to be asking than "what will this get me," and it's a habit that's harder to build once a kid's already spent years optimizing for the number.