What is a Control Group?

A control group is the set of participants who do not receive the thing being tested, giving researchers a baseline for comparison. Without one, there is often no way to tell whether the treatment caused the change or whether it would have happened anyway. It is what separates a study from a story.

The problem a control group solves is that things change on their own: people recover from colds untreated, students improve over a school year, sales rise every December. So a group that got the treatment and then improved has proved very little by itself. Run a second group treated identically except for the one thing being tested, and now you have a comparison, because whatever happens in that group is what would have happened anyway. Who goes into which group has to be decided at random, since if people choose for themselves the keenest and healthiest will pick the treatment. In medicine the control group is usually given a placebo, a dummy pill with no active ingredient, because people who believe they are being treated often do feel better. Neither the patients nor their doctors are told who got the real drug, so belief cannot do the drug's job. Some questions cannot be tested this way at all, since nobody can randomly assign people to smoke. Without a proper control group, researchers must be much more cautious about claiming that one thing caused another.

Baking the same recipe twice, changing only the oven temperature. Swap the flour too, use a different tin, bake on a humid day, and the second loaf will come out different with no way to know which change did it. The second loaf is not there to be eaten. It is there to be compared.