An unforeseen and unaccounted-for factor that generates an apparent association between two variables because it has a prior association with both. For example, height and school achievement may both be associated with socioeconomic status. Thus, an apparent link between size and success could be said to be subject to confounding by social class. Such a variable jeopardizes both the reliability and validity of a study.
See Causal pathway, Confound, Contamination, Experimental method, Reliability, Validity, Variable