Practical work: Its effectiveness in primary and secondary schools in England

30 practical lessons (10 primary, 20 secondary); 857 students observed

Non-selective English primary (n=10 schools) and secondary (n=20 schools) science classrooms, ages ~7-16.

Baseline (pre-training) arm of the evaluation of the national 'Getting Practical: Improving Practical Work in Science' (IPWiS) project. Multi-site case study with a condensed fieldwork strategy; audio-recorded discussions, interviews, observational field notes. NO control group, NO achievement test, NO effect size — the same two-level Millar/Tiberghien effectiveness framework as Abrahams & Millar 2008 (Level 1 = do what was intended; Level 2 = think what was intended). Coding done by the evaluators against teacher-stated intentions. All 30 observed tasks were highly structured 'recipe' tasks. The authors were the appointed evaluators of the IPWiS programme, which is why independence is recorded as developer-involved rather than independent — but note that this baseline arm reports on practice BEFORE the intervention, which blunts the usual developer-evaluation bias.

Extends the 2008 finding to primary schools and to a larger sample (30 lessons, 857 students): in BOTH primary and secondary science, 'recipe'-style practical work was highly effective at getting students to do what the teacher intended with the materials, and much less effective at getting them to use the intended scientific ideas to interpret what they had done. Neither primary nor secondary lesson plans contained explicit strategies for linking observation to concept. Primary tasks were shorter and spent more time on new scientific vocabulary. Crucially the paper also surveys the prior literature and reports that reviews of practical work (Hofstein & Lunetta 1982; Lazarowitz & Tamir 1994) have consistently concluded that, when outcomes are measured by pen-and-paper tests, practical work confers NO significant advantage in conceptual understanding over non-practical instruction. Again: no control arm, no effect size.

Not applicable: descriptive, no between-group comparison, no attainment outcome.

Replicates Abrahams & Millar 2008 in a larger and broader sample and extends it downward to primary schools. The two papers are the same team and the same analytic framework, so this is a conceptual extension by an overlapping author group, not an independent replication.