control group

control group A group used for comparison with another, either because it represents the most common or typical case, or because it illustrates the absence of some phenomenon being studied. Human beings are not inert, lifeless matter, and hence cannot be studied simply by carrying out before-and-after studies around some experimental stimulus applied to them. Many other developments and changes will be occurring spontaneously at the same time, as people go about their lives, and it is difficult to disentangle the effects of one particular stimulus from all the others affecting people, as individuals or as groups, at the same time. The solution is to identify a group, or social aggregate, that serves as a control group in providing information on characteristics or changes in a baseline situation that is as close to ‘ordinary’ as possible, or illustrates social behaviour in the absence of the key factor of interest. Control groups can be created at the sampling stage, through statistical manipulation of the data-set, or during analysis. The most rigorous type is experimental controls, obtained by random allocation of cases. See also CONTROL (EXPERIMENTAL).