One of the primary features of good research is that the experimenters attempted to control extraneous factors that might influence the study results. A control is anything included in the study that is designed to eliminate or minimize a potential threat to the validity of our conclusions. Controls can be very general and involve factors such as a believable description of the study purpose, careful attention to sampling through inclusion and exclusion criteria, good specification of constructs, and careful design of the laboratory setting. Controls can also be very specific to design strategies such as unobtrusive observation in naturalistic observation research, counterbalancing of experimental conditions in within-groups designs, and the inclusion of multiple control groups in between-groups designs

Why are controls necessary in psychological research? Rarely is behavior influenced by just one factor. Thus, in any research study, we may want to learn more about how a particular IV or predictor influences behavior but there may be other factors that are also operating that could explain the same behavior. By building controls into our research, we try to isolate the influence of our IV or predictor variable. If we fail to include adequate controls, we may not be able to explain our findings.

How do we anticipate in advance what controls might be necessary in our research? The best ways are by reading the literature and consulting with others. An exercise that can help us think about factors that may need to be to be controlled is to write the "Limitations" section of our research report in advance of conducting the studies. This exercise has helped many students identify and control factors that might influence study results.