CHP. 11 - Confounding vs. Obscuring Variables Flashcards
Confounding Variables
Variables that cause a false positive result
- threat to internal validity
Maturation Threat
Effect due to spontaneous changes in behavior
- due to passage of time
History Threat
Effect due to external events that impacts most members of a treatment group
Regression Threat
Threat due to “regression to the mean”
- aka regression to the average
Attrition Threat
May occur if there is a reduction in participants from pre- to post-test
- more of a problem if “only high” or “only low” scores drop
Testing Threat
Practice or fatigue effect
- Solution: use a post-test only design
Instrumentation Threat
- Instrument Decay, Observer Fatigue, 2 versions of a test aren’t sufficiently equivalent
- Solution: use a post-test only design
Testing vs. Instrumentation Threat
- Testing means that the participants change over time after being tested before
- Instrumentation means that the measuring instrument has changed over time
Selection-History Threat
Outside event affects people in a study but only 1 level of IV
Selection-Attrition Threat
Only 1 experimental group experiences attrition while the other doesn’t
Double Blind Study
Neither the participants nor the researchers know who is in the treatment group vs. comparison group
Single Blind Study
Participants know which group they are in but the observers don’t
Placebo Effect
Inert treatments can result in real improvements in symptoms b/c people believe that they’re receiving a valid treatment
Solutions for Confounding Variables
- Include comparison groups
- Use random assignment whenever possible
- Use double blind whenever possible
- Use reliable coding procedures
Obscuring Variables
Variables that can cause null results (miss)
