Forensic tests attempt to answer an important question. Usually it can be distilled down to: did he do it?
A DNA test attempts to answer this question by matching a biological sample to a specific person. A fingerprint comparison tries to connect a small unique pattern left at a crime scene to a person. A blood alcohol test endeavors to reveal if a blood sample contains ethanol. And if so, how much?
These tests are commonly done in duplicate. Duplicate testing means the tests were conducted by the same person, at the same time, using the same equipment. When tests are done in duplicate, and the results are close together, it’s easy to jump to the conclusion that the results are accurate.
That is, until you understand the difference between precision and accuracy.
Precision determines the closeness of two or more measurements. Take a person weighing 250 pounds. He steps on a bathroom scale. It reads 150 pounds. He steps off the scale.
In disbelief, and while smiling, he steps back on. Now the scale says 151 pounds. He is still smiling…because he knows that his weight is nowhere near either 150 pounds or 151 pounds.
The fact that these measurements are close together makes them precise. However, the closeness of the two measurements, does not demonstrate accuracy.
Accuracy is the closeness of the results to the target. Here, the true weight of the person (250 pounds) was the target. Accuracy is measured by how close the results were to the true value (i.e. 250 pounds). These results were approximately 100 pounds off. Not very accurate. However, they were very precise (only a pound apart).
Precision is a good sign of reliability. Unless it’s the only sign provided. Precision of forensic results is a necessary component in deciding trustworthiness… it’s just insufficient. It’s certainly not accuracy.
If the question you seek to answer requires accuracy, then precision won’t tell you: did he do it?