Validity and Reliability Flashcards
A threat to internal validity due to people dropping-out of the study.
e.g., People may quit, may be unable to locate, or die.
Experimental mortality
A threat to internal validity where people tend to improve on tests because of familiarity or practice with the test. e.g., If one practices taking SAT tests, one’s score is likely to improve.
Testing
A threat to internal validity…tendency to move closer to the average, or the natural regression toward the mean.
e.g., If really bad off, can only get better–move closer to the mean.
Statistical Regression
A threat to internal validity due to people continually growing & changing whether part of a study or not. e.g., A child may outgrow having accidents in the bed at night, having nothing to do with mom offering incentives to go to the toilet in the middle of the night.
Maturation/Passage of Time
A threat to internal validity when an outside event coincides in time with the manipulation of the independent variable. e.g., Students in research learn the definition of a research term, not because of the flashcards, but because it has been linked to a scandal reported widely on the news.
History
A threat to internal validity when measures may not be comparable. e.g., If the pre and post test are not exactly alike, one may be easier than the other. If the pre-test is easier, the post-test may not show improvement even if there was some
Instrumentation
The extent to which causal relationship depicted in the study can be generalized. Can one generalize findings from the sample to larger population?
External validity
When those being observed are aware of it (e.g., surveys, experiments, SSD).
Obtrusive Observation
A threat to internal validity where participant’s may not be truly comparable.
e.g., Dog lover may be more likely to participate in a survey about puppies.
Selection Bias
Those being observed are unaware of it (e.g., records, process notes, editorials).
Unobtrusive Observation
A threat to internal validity when you are not sure which happened first.
e.g., Does second-hand smoke cause lung cancer? Not everyone who works in a bar gets lung cancer.
Causal time-order
When several different research methods are used to collect data on the same phenomena.
e.g., conducting a online survey and agency records.
Triangulation
____ asks the question, “Did experiment stimulus really cause a change or effect the dependent variable?”
Internal Validity
When participants realize they are being studied, their behavior may change because they are being evaluated, not because of the intervention. _____ threatens the external validity.
Reactivity
