Experimenter bias and participant reactivity can threaten both internal and external validity.