Scientific modelling and experimental evidence
Physics connects simplified models with measurements whose precision is always limited.
1. Models and assumptions
A scientific model keeps the features needed for a question and deliberately ignores others. A particle model neglects size; an ideal string has no mass; constant acceleration ignores changes in resistance. A useful model states its assumptions, makes a testable prediction, and has a known range of validity. Disagreement with reliable data tells us to improve the model or the experiment.
2. Measurement and uncertainty
A measured value is reported as x = xbest ± Δx. The uncertainty is not a mistake: it describes a reasonable interval. For one analogue reading, half the smallest scale division is often a practical estimate. Repeated readings reduce random scatter; their mean is a better estimate, and half the range can estimate uncertainty. Relative uncertainty is Δx/x, often written as a percentage.
| Effect | Pattern | Response |
|---|---|---|
| Random error | Readings scatter above and below | Repeat, average, improve resolution |
| Systematic error | All readings shift similarly | Calibrate, check zero, change method |