Dragging a figure is how a student discovers what the theorem actually claims
Marton's variation theory holds that a critical feature of a concept becomes discernible precisely when it is the one thing held invariant while everything else varies. A dynamic figure makes this physical: grab a vertex, move it, and everything that can change does — but the property the theorem asserts stays fixed. The angle sum that refuses to leave 180° as you flatten the triangle is variation theory you can feel.
Research on dragging in dynamic geometry — Arzarello and colleagues' catalogue of dragging modalities, Baccaglini-Frank & Mariotti's "maintaining dragging" model, and Leung's work on the discernment of invariants in Educational Studies in Mathematics — finds that students who drag don't merely confirm a stated fact. They generate the conjecture themselves, arriving at a statement with a premise, a conclusion, and the conditional link between them.
The measured payoff is large. Chan & Leung's systematic review and meta-analysis (Journal of Educational Computing Research) reports a standardized mean difference around 1.0 for dynamic geometry over static instruction; a 2024 three-level meta-analysis (Ji, Guo & Song — 107 studies, 10,507 students) reports a moderate-to-strong effect of d ≈ 0.63.
- See draws the figure; Break hands over the drag handles. Dragging A, B, C while "∠A + ∠B + ∠C = 180°" stays put is the variation-theory contrast made literal — the invariant is the theorem.
- Put a prediction prompt before the drag ("what will the angle sum do as the triangle flattens?"). Prediction sharpens attention to the invariant that follows.
- Vary one thing at a time so the critical feature is separable from the noise — drag a single vertex, freeze the rest.