AcademyMeasurement and Vectors
Academy
How Physics Builds Models
Level 1 - Physics topic page in Measurement and Vectors.
Principle
A model predicts measured behavior by keeping only assumptions that matter.
Method
- Choose the system: the object or region being modeled.
- State the assumptions: what is kept, ignored, or held constant.
- Make a prediction that could be compared with a measurement.
- Test the residual against the measurement resolution.
Model residual
\[r=\text{measurement}-\text{prediction}\]
Model test
\[|r|\lesssim \text{uncertainty}\]
Checks
- Assumptions must be visible.
- Predictions need units.
- Residuals should not show a pattern.
- Useful does not mean exact.
Notes
- Add detail only when it improves a measurable prediction.
- A failed test can expose bad assumptions, bad data, or both.