# Precision

## What is precision?

For the SDC data to be taken seriously by scientists, it must be precise. Data precision is the degree to which repeated measurements reproduce a specific value. For example, 3.1415 is a more precise measurement than 3.1 because it goes out to many more decimal places.

The SDC can measure the mass of a dust particle to within a factor of two. This means that if a particle were measured with a mass of 3 picograms, scientists would report that the mass of the particle was somewhere between 1.5 picograms and 6 picograms.

## Precision versus accuracy

Precise measurements are not the same as accurate measurements. An accurate measurement is “correct” and agrees with scientific theory. So it is possible to have data that are accurate but not precise, or precise but not accurate. For example:

Let’s imagine you have a mechanical dog that runs a mile in exactly two minutes—never faster, never slower. You record your dog’s runs five times and your stopwatch reads 2.999 minutes every time. That means the stopwatch is extremely precise—it always gets the same measurement. However, the problem with your precise stopwatch is that it is not very accurate; you know that your dog runs a mile in exactly two minutes, not 2.999 minutes. So, you decide to get a more accurate stopwatch. You run your dog five more times, and get the following times with your new stopwatch: 2.000, 2.002, 2.003, 2.004, and 2.002 minutes. All of these values are more accurate than you got with the precise stopwatch, but they are less precise because the values are all different.

Why do precision and accuracy matter? Without accuracy, data are hard to interpret, but without precision, you can’t learn anything specific from them.

## Errors in precision

The main contribution to error in the precision of a measurement is random background noise. These tiny, random fluctuations in the detectors and electronics can wash out precise signals. Better electronics and instrument design can help minimize the amount of error but you can never eliminate it entirely.

The easiest way to improve precision is by developing a better instrument. As time passes and technology improves, scientists and engineers build on past experience to improve instrument design and construction; with time comes improved understanding and better technology.

Another way to improve precision is by taking more measurements. The more measurements scientists take, the more they can “average out” any random fluctuations. For example, let’s say that one person measures ten grains of rice and comes up with an average length; and another person measures 10,000 grains of rice and also comes up with an average length. Which person’s measurement will be more precise? The second person’s average will be more precise because any inaccurate measurements that he or she made will have less influence on the final average length.