Use the chain question → variables → method → assumptions → risk/ethics/heritage checks → repeatable evidence. Reproducible does not mean every result is identical; comparable results may still vary naturally.
Record materials, quantities, equipment, sequence, timing, units and conditions. Changing more than one relevant factor creates alternative explanations.
An analogous model represents selected features. State what it represents and omits: dialysis tubing, a shake table and a baking-soda volcano are useful only within their assumptions.
A hazard can cause harm; risk combines likelihood and consequence. Ethical planning includes consent, privacy, voluntary participation and appropriate use of biological or secondary data. Fieldwork must respect heritage sites, artefacts and protocols on Country/Place.