Accuracy vs Precision: What's the Difference in Science?
Accuracy is how close a measurement is to the true or accepted value. Precision is how consistent or repeatable a set of measurements is — how close they are to each other. A measurement can be precise but inaccurate (consistently wrong), accurate but imprecise (randomly scattered around the right answer), or both precise and accurate (consistently close to the true value).
In everyday language, "accurate" and "precise" are often used interchangeably — but in science, statistics, and engineering they have distinct and non-interchangeable meanings. Getting this distinction right matters in medicine (a precise but inaccurate diagnostic test is worse than it appears), in manufacturing (precise but inaccurate parts will all fail together), and in research (precise measurements that are systematically biased produce confidently wrong conclusions). The classic illustration is a dartboard: you can throw all your darts close together (precise) in the wrong place, or scatter them around the bullseye (accurate but imprecise), or cluster them at the bullseye (both).
Key Differences at a Glance
| Feature | Accuracy | Precision |
|---|---|---|
| Definition | Closeness to the true/accepted value | Consistency or repeatability of measurements |
| What it measures | How far from the correct answer | How spread out repeated measurements are |
| Error type | Related to systematic error (bias) | Related to random error (noise) |
| Can exist without the other? | Yes — accurate but scattered | Yes — precise but biased |
| Dartboard analogy | Darts near the bullseye | Darts clustered close together |
| Statistical measure | Closeness to true value (low mean error) | Low variance / standard deviation |
| Example | Scale reads 70.1 kg; actual = 70 kg | Scale always reads 65 kg for same person |
Accuracy: Getting the Right Answer
A measurement is accurate if it is close to the true or accepted reference value. Accuracy relates to systematic error — a constant bias that consistently pushes all measurements in one direction. If a thermometer reads 2°C too high regardless of temperature, it is systematically inaccurate. Accuracy is assessed by comparing measurements against a known standard or true value. In chemistry, a balance that consistently reads 0.5g above the true mass is inaccurate regardless of how repeatable its readings are. Improving accuracy requires identifying and removing systematic errors: calibrating instruments, correcting for known biases, using appropriate measurement methods.
Precision: Getting Consistent Answers
A set of measurements is precise if they cluster closely together — if repeated measurements of the same thing produce similar values. Precision relates to random error — the scatter around the average measurement. A precise instrument produces consistent, reproducible results. High precision is indicated by low standard deviation or variance. Note that precision says nothing about whether those consistent readings are correct — a precisely broken clock shows exactly the same (wrong) time every time it's read. Improving precision requires reducing random noise: using more sensitive instruments, controlling environmental variables, increasing the number of measurements.
The Four Combinations
There are four possible combinations of accuracy and precision. High accuracy, high precision: measurements cluster tightly around the true value — the ideal. High precision, low accuracy: measurements cluster tightly but consistently away from the true value (systematic bias). High accuracy, low precision: measurements scatter randomly but average close to the true value. Low accuracy, low precision: measurements scatter randomly far from the true value. The precision/accuracy distinction matters most when only one is present. A precise but inaccurate medical test will reliably produce wrong answers — a more dangerous failure mode than an imprecise but unbiased one.
Frequently Asked Questions
What is the difference between accuracy and precision?
Accuracy is how close a measurement is to the true or accepted value. Precision is how consistent or repeatable measurements are. You can have high precision (consistent results) that are inaccurate (all wrong by the same amount), or high accuracy (results average near the true value) but low precision (scattered results).
Can something be accurate but not precise?
Yes. If you measure a 100g weight multiple times and get readings of 98g, 103g, 97g, and 102g — the average is 100g (accurate), but the measurements are scattered (not precise). This happens with instruments subject to random noise.
In what fields does this distinction matter most?
Medicine (diagnostic tests), analytical chemistry, engineering and manufacturing (tolerances), scientific research (statistical analysis), navigation (GPS accuracy vs. precision), and statistics (bias vs. variance). In machine learning, the bias-variance tradeoff is directly related to the accuracy-precision distinction.
What is the difference between accuracy and precision in statistics?
In statistics, accuracy maps to low bias (the expected value of estimates is close to the true parameter), and precision maps to low variance (estimates cluster closely together). High-bias estimators are inaccurate; high-variance estimators are imprecise. The bias-variance tradeoff describes how reducing one often increases the other.
How We Write These Comparisons
SmartAss Facts comparisons are written to be the clearest, most accurate answer to "what is the difference between X and Y?" on the internet. We start from the primary definition — taxonomic, scientific, or linguistic — and work outward to the practical distinctions most people actually need.
Each comparison table row is independently sourced. If a distinction is more nuanced than a table cell allows, the detail appears in the body sections below the table. Last reviewed: 2026-05-25.
More Comparisons
Every comparison below is written to the same standard — primary sources, clear definitions, and no fudging the hard cases.