GIZMOS · RUNG IV · INNOVATE

They build the argument

Students cannot build a Gizmo. They can build the thing that uses one. This page is honest about how much room that leaves — less than the other tools in this course, and worth knowing before you plan around it.

Idea starters, not finished practice. None of this is an expectation. On whether the argument below is a good argument in your discipline, you are right and I am not.

Start with the honest part

The bar for Innovate in this course is strict: not faster, not easier — previously impossible. If a move is merely quicker than what you did before, it belongs on a lower rung and there is nothing wrong with that.

Held to that bar, Gizmos is thinner than Studio, Canva or Vivi, and it is thin for a structural reason. Students manipulate controls somebody else built. They cannot script, modify, extend or author a simulation. Teacher Presets — saving a configuration and reloading it — are teacher-only. There is no student-authoring surface in the product at all.

So the honest Innovate move here is not students make a Gizmo. It is students author the artifact that the Gizmo is evidence inside of. That is a real move and it is narrower than the rung usually is. Saying so is better than inventing something.

The two features this rung is actually built on

The screenshot tool

Students capture all or part of a Gizmo — including the graph and the data it generated — then copy or download the image and paste it into their own document. There are draggable annotation arrows for pointing at the thing they mean.

This is the student-facing feature that matters, and almost nobody uses it.

Using the screenshot feature →

The Data tab

Many Gizmos let a student export the numbers the simulation generated. Not a picture of a graph — the values, which they can then do something with.

A student who ran twenty trials has twenty rows nobody handed them. That is the raw material for a claim.

Every student control, including Data →

The move

THE ASSIGNMENT IS NOT THE GIZMO

Give a claim that is contested rather than a procedure that is known. Two students disagree about what happens when you change this. Design the runs that would settle it, run them, and make the case.

What the student hands in is theirs: the trials they chose, the screenshots they took, the arrows they placed, the data they exported and the paragraph that says what it shows and what it does not. The Gizmo appears inside it as evidence. It is not the assignment.

Where this clears the bar: the number of trials a student can run used to be rationed by apparatus and by the length of the block. A student got three runs because there were six spring scales and 45 minutes. Twenty runs, varied deliberately, one variable at a time, in four minutes — that was not available. Not faster: different in kind, because the argument a student can make from twenty trials is not the argument they can make from three.

Where it does not clear the bar, and this is worth saying: a poster made from Gizmo screenshots is a nicer version of a poster. If the deliverable does not require choosing which runs to do, it is Enhance wearing an Innovate costume.

Two objections I have not resolved

One. The simulation already knows the answer. A Gizmo models a relationship someone decided to model, correctly, in advance. A student who runs twenty trials inside it discovers the model, not the world. That is a genuinely different epistemic act from an experiment, and a science teacher may reasonably think the difference matters more than the volume of trials. I do not know how much it matters. You do.

Two. The artifact is where the AI question lives now. The moment the deliverable is a written case rather than a filled-in sheet, the honest-work question arrives with it. The screenshots and the exported data are actually an asset here — they are the student\u2019s own trials and they are hard to fabricate convincingly — but the paragraph around them is not. Worth pairing with Change the evidence, not the topic → rather than pretending the question is not on the table.

The cheapest version of this rung takes one class and no new tools: students run their trials, take two screenshots, and then say the argument out loud for 90 seconds instead of writing it. You get the reasoning, the AI question mostly evaporates, and you have graded it before the block ends. Every student speaking — oral evidence you can actually grade →

Where it breaks

Nothing comes back to Canvas, so the artifact has to. Gizmos has no grade passback. That is actually convenient here — the thing you are grading is a document or a recording the student submits to a Canvas assignment, so the gradebook works normally. Grade the case, not the simulation.

The screenshot tool is unknown to students. Nobody discovers it. Show it once on the room screen, in about 40 seconds, or it will not get used.

The accessibility floor is low. Some interactive Gizmos cannot be driven from a keyboard and there is no spoken audio. A student who cannot operate the controls can still author the argument — run the trials they specify, hand them the screenshots, and grade the reasoning. The reasoning was always the point.

Next

Rung II · Enhance

The prediction gate. Cheaper than this page and it does more.

Predict before you drag →

Rung III · Deepen

Read the Heatmap before you plan the next class, and let it pick the warm-up.

The Heatmap picks tomorrow →

Back to the hub

Setup, the four sheets, the three products, and what will bite you.

Gizmos — the simulation is the easy part →

Students cannot build the simulation. They can decide which twenty runs would settle the question — and that decision was rationed by apparatus until about ten years ago.

Jason Beyer · Educational Technology · Regis Jesuit High School · A.M.D.G.

HelpDesk@regisjesuit.com · 303.269.8080 · LAST VERIFIED — AUGUST 2026