What is the Difference between Accuracy and Precision Measurements?

According to a thesaurus, precision is a synonym for accuracy. However, the scientific world distinguishes between these two concepts in important ways. In fact, different fields use slightly different terminology at times (in statistics, error, bias, and variability are terms more commonly encountered), but we are going to focus on a general overview of what these terms mean and how they are used.

What is Accuracy?

Accuracy refers to how well something approximates a real, or true, value. Let’s use a couple examples to explain this more clearly.

Example 1. The common example to illustrate this concept is a dartboard. Darts situated within the bullseye are accurate, because they are located in the correct position with respect to the rules of darts.

Example 2. If you design an IQ test to measure intelligence, how well that test can measure intelligence (and therefore approach a representation of the real intelligence of a person who takes the test) is the accuracy of that test.

What is Precision?

While accuracy points to the correctness of a measure, precision refers to measures compared to each other. Something is precise if the measures are close to each other in value.

Example 1. On a dartboard, the darts are precise if they are close to each other on the board. It does not matter if they are close to the bullseye, because precision only refers to measures in comparison to each other. This means the darts might be clustered together at the bottom of the dartboard, yet they are still precise because they are close to one another!

Example 2. For the IQ test, it can be said to be precise if people of similar aptitudes get similar scores, or if someone who takes the IQ test multiple times scores around the same each time. Note that the IQ test in this case does not have to measure intelligence, it only has to produce similar results. It could, for instance, be measuring general knowledge about history!

So How Whats The Difference Between Accuracy and Precision

For a measurement to be considered scientifically valid, it must be both accurate and precise. This is to say that the measurement has to accurately reflect a true, or accepted, value, and it also has to do so reliably (precision is reliable, or reproducible, because it generates the same or a very similar result).

It is important to clarify that a precise measurement is not one that always clusters values together. You can have a precise IQ test in the sense that people with similar skills score similarly, or if one person takes the test over and over, they continually receive an identical result.

However, the clusters of scores will vary based on the ability of the test takers. There will be people who score high, others in the middle, and others at the low end. To be precise, the measure needs to have similar test takers scoring close to one another.

Dartboard Example

Darts board are often used to explain the difference between accuracy and precision.

At the point when Accurately hitting an objective this implies you are near the focal point of the objective, ccurately hitting the target means you are close to the centre of the target, even if all the marks are on different sides of the centre. Precisely hitting a target means all the hits are closely spaced, even if they are very far from the centre of the target.

More Examples for Better Understanding

Finally, let’s go over some examples of how something can be accurate but not precise, both, neither, or precise but not accurate.

Example 1. Accurate but not precise. A measure might be accurate in that it approximates the correct value, but does so with a lot of variability. Going back to our IQ test example, this would mean that the scores do approach actual intelligence values, but there are wide differences amongst test takers with similar aptitude who should have scores that are closer in value due to their similar aptitudes.

Example 2. Precise but not accurate. A measure will replicate results over and over, but they are not correct. Remember that precision does not rely on truthfulness, and the reference made earlier to an IQ test measuring history knowledge instead of general intelligence. In this sense, test takers with in-depth knowledge of history should all score about the same, indicating the test has high precision. However, it is not measuring what the test intended to and is thus inaccurate.

Example 3. Neither. If a measure is not accurate or precise, the scores will be all over the place and will not reflect correct values. In the IQ test, this would mean both high and low ability test takers could score high, low or somewhere in between. The test will not represent their actual general intelligence scores. Further, if two people of similar abilities take the test, one might score very high and the other very low.

Example 4. Both. If a measure is both accurate and precise, scores will reflect actual correct values and test takers of similar ability will have scores that are close in value to each other. This is the ideal situation, because the measure is able to produce correct values repeatedly, making it a trustworthy instrument for whatever field it is being used for.