As many people probably don’t know, there is an actual difference between accuracy and precision. In an experiment, to be precise means that all the measurements you collect have almost the exact same value. For example, if you were measuring an amount of water, and you gathered masses of 700.075 grams, 700.076 grams, and 700.075 grams, all your measurements are pretty precise compared to 700.075 grams, 750.0 grams, and 8 grams. High levels of precision indicate that an experiment can be repeated by others and the results can be reproducible, which is often very good.

Accuracy, on the other hand, is determined by how close your measurements were to the true value, which is often a defined number or is given. For example, imagine you had 500 cubic centimeters of water, and measured the volume three times. Being accurate would be having measurements of 500 grams, 499.999 grams, and 500.001 grams, because they are all very close to the true mass of 500cc of water (which is 500 grams, because 1cc of water= 1 gram of water).

Also relating to accuracy and precision is the concept of significant digits. Significant digits are the numbers in a measurement that are useful. They are counted following a strange rule, called the “Atlantic-Pacific” rule by our chemistry teacher, at least. If a decimal is absent, you count from the Atlantic side (right) of the number, the first non-zero number and every number after it. If a decimal is present, you count from the Pacific side (left), the first non-zero number and every number after it. For example, in a speed of 200.5 feet per second, there are four significant digits: the 2, 5, and the two zeros. In 300 oz, there is only one significant digit, the 3, because the two zeros carry uncertainty; they may have been rounded for all we know. In 0.0860 tons, there are three “sig dig’s”: the 8, 6, and the last 0.

Significant digits are important because they are directly related to precision. The more significant digits you have, the more precise you are, and so also the more repeatable your experiments are. For example, if you went about calculating the value of π (pi), then 3.141592653589 would be much more precise (and accurate) than just 3.14, because the first one is over 592-million trillionths closer to π’s true value.

Advertisements