Common Threats to Internal Validity

Question	Answer
Mrs. Haiti	Maturation Regression Selection History Attrition Instrumentation Testing Effects Interaction
Maturation	Change that occurs within subjects that may just be due to presence of time (Longitudinal - but not only)
Regression	Tendency for extreme scores to move (regress) closer to the mean or average score.
Selection	Who are the subjects? Are the comparison groups comparable? Minimize differences between groups using random sampling. What if you cannot randomly assign? Then it is not a 'pure' experiment (quasi-experiment)
History	Events in subjects life/environment, other than manipulated IV that occurs during the experiment and may effect outcome.
Attrition	Loss of subjects. Not a problem if even proportion are lost between groups (e.g. Differential attrition). Loss of subjects within subjects - problematic if there is something 'different' about the people that are leaving the study.
Instrumentation	Unwanted changes in characteristics of instrument/procedure
Testing Effects	e.g. The Hawthorne Effect
Interaction	Combination of any number of threats to internal validity.

Resource summary

	Created by Kyle Simpson over 8 years ago