Fostering students' knowledge and argumentation skills through dilemmas in human genetics

Experimental and comparison classes of ninth-grade biology students in Israel (exact n not obtained)

Israel; grade 9 (~age 14-15) biology students studying human genetics.

The best-cited demonstration that argumentation can be taught inside science content. The treatment is a genetics unit with explicit, direct instruction in general argumentation structure — this is NOT open inquiry or discovery; students are told what a good argument is and given dilemmas to practise on. Comparison classes studied the same genetics content without the explicit argumentation strand. Not randomised. Both outcome instruments were built by the authors: a genetics knowledge test and a coded argumentation task. The "transfer" claim is the load-bearing one, and it is transfer from genetics dilemmas to everyday-life dilemmas — a structurally identical argumentation task with a different cover story, scored by the same rubric by the same team. That is near transfer under this archive's taxonomy no matter what the paper calls it. Nothing here measures attainment in another subject, a general reasoning test, or any independent standardised instrument.

Explicit teaching of argumentation inside a genetics unit improved both argumentation quality and genetics knowledge: the experimental group scored significantly higher than the comparison group on a genetics knowledge test, and more students cited correct specific biological knowledge in their arguments (only 16.2% did so before instruction, while ~90% could already produce a simple argument). Students also applied the taught reasoning to everyday-life dilemmas. The honest reading is that a reasoning frame taught explicitly transfers across cover stories within the same task type and improves content learning as a by-product — a domain-skill and near-transfer result, not evidence that school science builds general reasoning.

Medium: non-randomised intact classes, so class composition and teacher quality are not controlled; both outcomes are author-designed and author-scored.

The core "explicit argumentation instruction improves argument quality" finding recurs throughout the argumentation literature. What did not replicate is the leap to student outcomes at scale: Osborne et al. (2013) put argumentation into four whole science departments for two years and found few significant changes against a comparison sample on standard instruments.