Schema abstraction with productive failure and analogical comparison
Among 96 ninth-grade students, comparing climate models with models from other domains improved far transfer. Both productive-failure classes learned.

The idea
A schema is useful when it travels into a domain the student has not studied. This classroom study taught complex systems and climate change with agent-based models under two productive-failure designs. In both, students explored a problem before teacher-led instruction. One group used climate models only. The other compared those climate models with complexity models from other domains.
The approach
The classes were ninth grade at a high-achieving Australian girls' school: 114 students enrolled, 96 in the final analysis after missing tests were removed. Both groups gained on declarative knowledge, explanatory knowledge, and within-domain near transfer. The two-model group, which had to compare across domains, scored higher on an across-domain far-transfer problem.
- 114 ninth-grade students enrolled, 96 in the final analysis
- Both designs start with productive failure, then teacher instruction
- One climate model, or climate models compared with models from other domains
- Both groups gained on knowledge and near transfer
- The comparison group scored higher on far transfer
- Jacobson, Goldwater, Markauskaite, Lai, Kapur, Roberts, and Hilton. Learning and Instruction, 65, 101222
The paper: hdl.handle.net .