Control (Scientific)

A control, or control group is a tool used to confirm whether changes are actually having an effect. The control group is exposed to the same conditions as the test group with the exception of the variable that is being examined.

For example, you may be experimenting whether increasing tire pressure makes a difference in mileage. You would keep all other factors the same (route, vehicle type, brand of gas, etc.) the same for two groups. The only difference would be in tire pressure. The control group would have the current tire pressure, and the test group would have the higher pressure.

Control groups are part of scientific rigor. They are not always necessary but can be beneficial to use when testing new processes or system. This is most often the case when you are unsure of whether a change will actually work or not, and there is a high risk or large cost of making an incorrect conclusion.

The control group, whenever possible, should be looked at during the same time period as the test group. This prevents other changes that you may not be aware of from affecting the results.

You may also use blind control groups. In the tire pressure experiment, you would not tell the drivers which group their vehicle falls into. This prevents them from biasing the results. In double blind tests, neither the tester nor the test subject knows which group they are in. This is most common in medical drug testing in which both the doctor and patient are in the dark as to whether they are using the actual drug or a placebo.