OUTLINE OF AN EXPERIMENT
(or How to Kick Butt on the AP Psychology FRQ)
I. Identify your subjects
- Provide a reasonable number (ex: 100-300 subjects AT MOST)
- Provide any subject characteristics that are important (ex: 100 subjects
suffering from depression)
- While it often is not possible, talk about selecting a random sample (or
representative sample) of subjects who AGREE to participate in your study.
- RANDOMLY ASSIGN your subjects to a:
- CONTROL GROUP and an
- EXPERIMENTAL GROUP
- Mention that random assignment will make the groups equivalent with
respect to PRE-EXISTING SUBJECT VARIABLES.
- Identify the:
- Independent Variable (the variable you will manipulate and that you
believe will cause a change in the dependent variable
- Dependent Variable (the variable you will measure and that you believe
will be AFFECTED by the I.V.)
IV. Operationalize (provide an
operational definition for) the:
- Independent Variable: Explain how you will MANIPULATE it. Explain how
the two groups will be treated DIFFERENTLY with respect to the I.V.
- Dependent Variable: Explain SPECIFICALLY how you will MEASURE the D.V.
- Discuss CONTROL techniques you will use and the CONFOUNDING VARIABLES they will eliminate. ALWAYS
include:
- Random Assignment
to the control and experimental groups.
This controls pre-existing subject variables
B. Use of a control group.
This controls for history, maturation, and testing effects
- Other Control Techniques: (Discuss these when they are appropriate for the
experiment)
- Single or double-blind design to control subject and experimenter biases
- Use of a placebo to control the placebo effect (which includes
experimenter and subject biases)
- Any other EXTRANEOUS variables you want to be the SAME in both groups.
This allows you to rule out alternative explanations for your results.
- Describe how you would EVALUATE your results
- Explain that you will COMPARE the control and experimental groups
with respect to the DEPENDENT VARIABLE
- Explain that you need to find a STATISTICALLY SIGNIFICANT DIFFERENCE
(not a correlation) between the scores of the two groups.
- Explain that to be sure of your results, you would probably want to REPLICATE
your experiment.
|