Calibration Fundamentals
Calibration links an instrument's reading to a known standard. Without it, all measurements remain systematically shifted.
| Error Type | Description | Example |
|---|---|---|
| Zero offset | Doesn't read zero at zero | Scale reads empty |
| Scale factor | Consistently high or low | Ruler expanded by heat |
| Non-linearity | Error varies across range | Spring past Hooke's law |
| Hysteresis | Depends on approach direction | Mechanical backlash |
Calibration procedure measures known standards across your range, plots measured versus true values, and fits a calibration curve (). Apply the inverse to correct future readings.
Errors & Accuracy
Detect systematic errors by comparing with accepted values, using different methods, or checking residuals (measured minus fitted values).
| Term | Definition |
|---|---|
| Accurate | Close to true value; low systematic error |
| Precise | Measurements cluster tightly; low random error |
Common pitfall: Averaging cures only random error. A miscalibrated instrument delivers precision without accuracy. Systematic errors do not decrease with more measurements. The only cure is identifying and removing the source.