Accuracy, Precision, and Correctness – What Do They Mean?
Three terms that are often confused — but describe very different things. Understanding them helps you evaluate measurement results much better.
Why the Difference Matters
When a measurement result differs from what you expected, the question immediately arises: Is it the device? The reagents? My technique? To answer that, you need three concepts — correctness, precision, and accuracy.
They sound similar, but mean very different things.
Correctness
Correctness describes how close a measured value is to the true value.
Example: You measure a reference solution with a known KH value of 7.0 °dKH. Your device shows 6.9 °dKH. The deviation is small — the device is correct.
Correctness can only be assessed when you know a true reference value — for example, through a certified reference solution or an ICP analysis.
Precision
Precision describes how well measurements can be repeated — regardless of whether they are correct.
Example: You measure the same sample ten times. All values fall between 6.9 and 7.1 °dKH. That is precise.
A device can be very precise but still incorrect — for example if it is systematically off by 0.5 °dKH, but does so consistently every time.
Precision without correctness: A broken thermometer always shows 20 °C — precise, but not correct. Correctness without precision: Each measurement randomly hits the true value, but the scatter is large.
Accuracy
Accuracy is the combination of both: an accurate measuring device is both correct and precise.
For reef keeping, a precision of ±1% relative to the true value is perfectly sufficient. At a KH of 8.0 °dKH, that corresponds to a range of 7.92 to 8.08 °dKH.
Resolution
One more concept: resolution is the smallest readable interval — like the millimeter markings on a ruler. High resolution does not automatically mean high accuracy.
Example: A digital scale shows three decimal places — but if it systematically reads 5 g too high, it has high resolution but is not correct.
For practical aquarium use, precision matters most: if your readings are consistent, you can reliably detect trends — even if the absolute value is slightly off. The fine adjustment in the Typus app corrects systematic deviations.