Internal validity refers to the extent to which a study can demonstrate a causal relationship between variables, free from confounding factors. It ensures that the observed effects in an experiment are attributable to the manipulation of the independent variable rather than other extraneous variables.