Your task is to design a social science experiment on a topic of your choice. Be sure to propose an
experiment that conforms to the “true” or “classic” experiment design so that you can randomly assign
subjects to their respective groups. DO NOT propose or design one of the quasi-experiments. You will
need to know about quasi-experiments for the final exam, but in the interests of making things easy for
you, I don’t want you to mess with them here. Completing the assignment will involve several important
steps:
1. You must decide exactly what it is that you want to study. I don’t care what you choose to
examine, as long as the variables are relevant to social science research. No medical
experiments. No experiments on health and fitness. Etc.
2. You will need to determine what the dependent and independent variables are, and describe in
detail how you will operationalize them. Then you need to defend these operationalizations and
convince me that they are valid! Recall that in experimental research we use different words for
these variables. For instance, the independent variable is often referred to as the “treatment.”
3. If there is a pre-test, you will need to describe it in detail, and state how it measures the
dependent variable.
4. You will need to describe how the treatment will be administered.
5. Finally, you must describe the post-test, and talk about how your research has identified a
causal mechanism. (Review the criteria for a causal relationship from chapter 6.)
6. You must assess how your experiment deals with the various threats to validity that often
plague experiments. These are found on pages 214 – 218. How can you best ensure that your
experiment is valid?
In short, I want you to conceive and design an experiment, and explain exactly how you would
implement it. You will be graded on how well you understand the logic of experiment design, and the
clarity and accuracy of your explanation of your experiment. Most papers will be around 3 pages in
length.
