Which type of investigation has a control group?

Medical type investigations require a control group; there are other types as well which may require a control group. The idea is to limit that number of variables. When you are trying out a new drug, you also have a group that has the same medical condition (Arthritis, whatever it may be) but you give them a placebo, generally a sugar pill, instead of the drug that you are testing. That way, in case there is some effect merely from the fact that people are taking a pill, you can find out the extent to which the actual chemical composition of the pill is contributing to the outcome.