Home | Company |  Products | Clients | News | Resources | Contact 

The dart board below illustrates both precision and bias.

Precision deals with how tightly clustered together the results of several samples are in relation to an expected value. If a particular sampling methodology produces very similar results every time it is used, then that methodology is precise. Bias deals with how close the values that emerge from samples approximate the true value. From this information, we can determine the following:

The Green values are both unbiased and precise. They are all near the “bulls-eye,” (true value) and relatively close together. This is an optimal sampling procedure.

The Blue values are unbiased, but not very precise. They are near the bulls-eye, but there is substantial variability in their locations.

The Red values are precise, but biased. Something causes that sampling methodology to deviate from the true value, but the methodology produces consistent results.

The Yellow values are both imprecise and biased. They are off the mark and also demonstrate high variability. 

Remember, in the real world, the location of the “bulls eye,” or true value, is unknown. We are trying to estimate it. To understand why some kinds of samples are better than others, imagine the dart board above with no bulls eye drawn in. Which samples would make estimating it the easiest? Which would be misleading?

Back to the glossary