Designing, Conudcting, and Analyzing Field Experiments, ICPSR Three-Day Course
Donald Green, Yale University
Designing, Conducting, and Analyzing Field Experiments
This short course aims to accomplish the following: (1) Explain why experiments are valuable tools for social science and program evaluation; (2) Examine in-depth examples of how field experiments are designed, executed, and analyzed; and (3) Explore and develop research ideas through discussion with peers and specialists. Experiments enable social scientists to draw valid inferences about cause and effect. The essential ingredient of experimentation is random assignment of people to treatment and control groups. Randomization ensures that these groups differ solely due to chance. So long as the experiment involves an ample number of subjects, the role of chance becomes minimal; the treatment and control conditions become virtually identical. These equivalent groups are then presented with different treatments. Since pre-existing differences have been eliminated, the different responses of the treatment and control groups may be attributed to the influence of the treatment. Experiments correct many of the deficiencies of observational, or nonexperimental, data. Random assignment enables researchers to disentangle the complex causal interplay among variables. It also affords the researcher much more control over what that treatment is and how accurately it is measured. The workshop will concentrate on “field” experiments, which take place in real-world settings: political campaigns, schools, police agencies, and the like. Participants should have a background in statistics, up to and including multiple regression and analysis of variance.
Fee: $1000 members; $1000 non-members