The key to experimental design
At the end of the DVs it is good to have a manipulation check
- These often give away the purpose of the study,
so do them last!
Experimental Realism
Does the manipulation have an impact on the participants?
Are participants involved in the research paradigm/situation?
Internal Validity
Confidence that the manipulation of the IV is responsible for the changes in the DV(s)
Experimental control and standardization of conditions lead to high internal validity
There are a variety of threats to internal validity...
Confounds
Unmeasured variables that vary systematically across conditions
Avoid by: random assignment to conditions whenever possible
- Never run an entire condition at once!!!
Placebo Effects
Avoid by: maintaining standardized expectations among P's across conditions
Examples:
- Alcohol and attractiveness
- Herbal pills and memory
- Experimental medication and clinical trials
Demand Characteristics
When participants can guess the research hypothesis
Avoid by: a good cover story, nonreactive measures
Experimenter Bias
Problems with self-fulfilling prophecies, biased perceptions, and even flat-out cheating
Avoid by: naïve or blind experimenters
Maintaining Internal Validity
random assignment to conditions!!!
standardization across conditions
pilot testing
deception
naïve/blind/mechanical experimenters
repeated-measures designs with baseline
Back to Overheads Page
Back to 381-002 Main Page