APA Dictionary of Psychology

a measure that is not under investigation in an experiment but may potentially affect the outcome or dependent variable and thus may influence results. Such potential influence often requires that an extraneous variable be controlled during research. See also confound.