Validity
MCAT glossary · Psychological, Social & Biological Foundations of Behavior
Definition
The degree to which a measurement or study captures what it claims to measure. Subtypes: INTERNAL validity (does the study design support a causal inference about the intervention's effect, controlling for confounders?); EXTERNAL validity (do findings generalize beyond the study sample and setting?); CONSTRUCT validity (does the operationalization actually measure the theoretical construct?); STATISTICAL validity (are the statistical inferences justified by the test assumptions and sample?).
Validity in plain English
Validity asks the honest question: is your study actually measuring what you claim, or something else? Imagine a bathroom scale — if it reads your weight but always adds 10 pounds, it's not valid. Researchers break this into flavors: INTERNAL validity asks whether your experiment truly shows that A caused B, rather than some sneaky third factor doing it; EXTERNAL validity asks whether your result holds for people and places beyond the specific group you tested; CONSTRUCT validity asks whether your measurement really captures the abstract idea you care about (does a survey of 'happiness' actually measure happiness?); and STATISTICAL validity asks whether your math and sample were solid enough to trust the conclusion. A study can nail one kind and fail another, so you have to check each.
Learn these first
Sign up free: all 171 MCAT terms as flashcards, plus practice questions with full explanations →
Related MCAT terms
MCAT practice questions · MCAT flashcards · MCAT study guide