08_Internal Validity and Confounds

Question 1

Internal validity refers to the ___

Answer 1

extent to which a research study is designed so that testing the hypothesis is unambiguous.

Question 2

what is Spurious relationship?

Answer 2

two or more variables are correlated, but the relationship is NOT causal.

Question 3

What are Lurking Variables?

Answer 3

“Interferes” in the relationship between two variables, X & Y.
   Lurking Variable 
Extraneous Variable 
Confound Variable 
Spurious Variable 
Third Variable

Question 4

Sources of Lurking Variables

Answer 4

Random Selection ≠ Random Assignment

Question 5

What are Non-Comparable Groups?

Answer 5

• Selection bias
Solution: Random Assignment (experiments) OR
Solution: Pre-test (non-experiments)

Mortality (the highest scores drop)
Solution: Drop matched pair, increase sample size

Question 6

What are Endogenous Change in Participants?

Answer 6

• Maturation (e.g. ageing)
• Testing (e.g. improvement happens because of retesting)
• Regression ( e.g. bottom 20%) on average have to do better

Solution: Experimental Design
Or
Solution: Repeated measures, time-series

Question 7

What is Treatment Misidentification?

Answer 7

• Experimenter Expectancies (self fulling prophecy)
• Placebo Effect (can cause an effect) 2/3 patients
• Hawthorne Effect ( research improvements/ working harder)
• Demand Characteristics (an expectancy effect)

Solution: Double-blind procedures
OR
Solution: Process Analysis (Fidelity Check)

Question 8

What is Contamination?

Answer 8

• One participant group influences another participant group.
  Solution: Lab setting, no contact between groups

Question 9

What are History effects?

Answer 9

• Historical changes in the environment occur during a study that influence the outcomes.

Solution: Make note of events, possibly exclude participants (“outliers”)

Question 10

What is Experimental Design?

Answer 10

Good at protecting internal validity; but not always good at protecting external validity

Question 11

What is Quasi-Experimental Design?

Answer 11

Good at protecting external validity; but not always good at protecting internal validity

Question 12

The main goal of ensuring good internal validity is ___

Answer 12

to eliminate the alternative explanations of research findings