Skip to main content
    Path to Science

    Thinking Dictionary

    Accuracy vs precision

    What it is

    When you measure something, two different questions matter.

    • Accuracy is how close a measurement is to the true value.
    • Precision is how close repeated measurements are to each other.¹˒²

    Picture a bathroom scale. One scale always shows 2 kg too much. It gives the same reading every time, so it is precise, but it is not accurate. Another scale jumps around a bit, but its readings centre on your real weight. On average it is right, but it is not precise. A good scale is both: it gives the same reading each time, and that reading is right.

    Experts in measurement add one more word. Trueness is how close the average of many measurements is to the true value. In their language, a single result is accurate only when it has both trueness and precision.¹˒² For everyday purposes, "accurate means on target, precise means consistent" works well.

    Why it matters

    We trust numbers. A number shown on a screen feels exact. But a reading can be consistent and still be wrong.

    The kind of error matters too. Some errors push every reading the same way, like the scale that always adds 2 kg. This is called a systematic error, or bias. Other errors scatter readings randomly up and down. Repeating a measurement and taking the average helps with random scatter. It does nothing about a systematic error. If every reading is 2 kg too high, the average is 2 kg too high as well.

    The way to find a systematic error is calibration: checking your instrument against something whose true value you already know. That known reference is called a standard.¹

    In everyday life

    • Fitness trackers. In one lab study, researchers tested seven wrist-worn fitness trackers. Most measured heart rate fairly well. For six of the seven, the typical error was under 5% while people were cycling. But none of them estimated calories burned to within 20%. The errors were also larger for people with darker skin and for people with a higher body weight relative to height.³ A tracker can give you neat, steady calorie numbers that are still well off.
    • Lots of digits. A thermometer that reads 37.46 °C looks more trustworthy than one that reads 37 °C. But the number of digits on a display only shows how finely it can show a number. It says nothing about whether the number is right.¹
    • Kitchen scales and measuring cups. A kitchen scale that always reads a little light will give you the same wrong answer every time. Your recipe will come out the same, but not the way it was written.
    • Checking a thermometer. Put a kitchen thermometer in a glass of ice water. It should read about 0 °C. If it doesn't, that's a calibration error, and you can't fully trust it on unknown temperatures.¹

    The common trap

    The trap is thinking that consistent means correct. If your device gives the same number again and again, you only know it is precise. You don't yet know it is accurate.

    Some ways out:

    • Check against something known. Before trusting a new device, test it on something whose true value you already know.
    • Don't be fooled by digits. More decimal places are not more accuracy.
    • Know what averaging can and can't do. Averaging smooths out random scatter, but it can't remove a steady bias.

    Try it

    You need your bathroom scale and something that weighs a known amount. An unopened 1 kg bag of sugar works, or 1 litre of water, which weighs about 1 kg. If your scale can't register such a light object, hold it and weigh yourself with and without it, then take the difference.

    Weigh the known object five times, taking it off the scale between readings. Write down each number.

    • How spread out are the readings? That shows the scale's precision.
    • How far is their average from 1 kg? That shows its bias, or lack of trueness.

    Now move the scale from a hard floor onto carpet and do it again. Does anything change?

    Related concepts

    Senses vs instruments · Multiple lines of evidence · Iteration · Credible sources · Fermi estimation · Random vs systematic error · Error bars and confidence intervals

    Sources

    1. Joint Committee for Guides in Metrology. (2012). International vocabulary of metrology: Basic and general concepts and associated terms (VIM) (JCGM 200:2012, 3rd ed.). BIPM. https://www.bipm.org/documents/20126/2071204/JCGM_200_2012.pdf

    2. Menditto, A., Patriarca, M., & Magnusson, B. (2007). Understanding the meaning of accuracy, trueness and precision. Accreditation and Quality Assurance, 12(1), 45–47. https://doi.org/10.1007/s00769-006-0191-z

    3. Shcherbina, A., Mattsson, C. M., Waggott, D., Salisbury, H., Christle, J. W., Hastie, T., Wheeler, M. T., & Ashley, E. A. (2017). Accuracy in wrist-worn, sensor-based measurements of heart rate and energy expenditure in a diverse cohort. Journal of Personalized Medicine, 7(2), 3. https://doi.org/10.3390/jpm7020003