Accuracy vs Precision
Accuracy and precision are similar concepts with a significant difference. Accuracy has to do with whether something is correct, precision has to do with how refined a measurement, calculation, or specification is. For example, I can state that the height of the Washington Monument is 305 feet 1 inch, which is quite precise. Unfortunately, the measurement is not at all accurate because this is the height of the Statue of Liberty. The height of the monument is 554 feet 7 11/32 inches.
The focus should be accuracy first, precision after. The level of precision should also be useful, not so crude as to prevent proper planning or execution and not so fine as to be unnecessarily exact, or worse, convey a level of certainty that is not justified by the data or process.
Application
-
To evaluate if something is accurate, consider the source and the process by which some value was determined.
-
To evaluate if something is accurate, consider the proposition that something else, even the opposite, is accurate.
-
If something is positioned as accurate but allows for judgement, ask (or provide) an associated level of confidence.
-
Be careful not to assume something is accurate simply because it is expressed with great precision. Similarly, don’t misrepresent your level of confidence that something is accurate by providing more precision than is justified.
-
If something is less precise than you want, ask or provide what it will cost to gain greater precision.
The terms ‘significant digits’ or ‘significant figures’ may get thrown around. In math, these are the number of digits used to express something to some desired level of precision. For example, 5.12 has three significant figures.