In the scientific method, each experiment should have two groups:
- The Experimental Group: this is the group on which a new variable is added
- The Control Group: this is the group where nothing changes – it is missing the variable introduced to the experimental group
The goal is that the variable being tested is the only difference between the two groups, so effects on the experimental group when compared to the control group can be attributed solely to the introduction of the variable.
A problem is bias for both a member of one of the groups, and for a researchers running the experiment:
- A member might alter their behavior if they know which group they’re in. If they know they’re receiving medication which they believe will lower their cholesterol (that being the purpose of the study), they might eat fattier food, thinking they now have protection from it. Or they might eat healthier food to try and amplify the effect of the medication.
- A researcher might alter their perceptions and evaluations of any single member if they know in advance which group the member is in. If they know a particular member is in the experimental group, they might be biased to see changes that don’t exist, and vice-versa.
In a “blind” or a “single blind” study, no member knows which group they’re in. If I’m a member of one of the two groups, then I have no idea if I got the variable or not, so I don’t change my behavior in any way.
The next step is a “double blind” study, where the members of the groups don’t know which one they’re in, and neither do the researchers. So, during the study, when evaluating any member, the researcher has no idea if this particular member was in the experimental or control groups, and therefore cannot be biased in their evaluation.
Presumably, the membership of the members is only revealed (to either side) when the study is complete.