Using comparison groups to rule out alternative explanations
Example: therapy outcome research
What possible confounds need to control for?
*Depends on research area
*What is state of the art?
*e.g., alcohol expectancy in MacDonald et al.
*Placebos in drug research
*Example: therapy outcome research
Therapy outcome research
*When researchers develop a new treatment for a psychological disorder, they want to
test the efficacy of the new treatment.
*3 types of confounds:
*Other threats to internal validity
*By-product of operationalization
“non-specific factors”— e.g.,
*therapist attention, warmth, & empathy
*ability to arouse hope in clients
*simply undergoing a procedure designed to ameliorate a problem
Common to all therapies