COM 381 Lecture Notes - Lecture 8: Causal Inference, Treatment And Control Groups, Statistical Significance
Document Summary
They are one of many methodologies available to social sciences. Science seeks causal explanation (better than association/relationship explanation) Causal: a change in one variable, it will cause a change in another variable. Association(relationship): if this one variable is present, then i can say with certainty that this other variable will be present. The logic of explanation: the basic features. Researcher exerts control over the independent variable. Value depends upon the manipulation of the independent variable. Group exposed to the manipulated independent variable. Only difference between the treatment and control groups is that the control group is not exposed to the manipulation. In an experiment, participants are assigned to groups at random to treatment and control groups. Change in one is associated with change in another, but not always with cause. Remove other explanations for the changes you see in the dependent variable. Test the significance - how likely it was that you would get that result.