The process of allocating individuals to groups (i.e., exposed and unexposed) by chance (Last, p. 150). If the sample size is large enough, randomization helps balance the distribution of known and unknown confounders between study groups (See Aschengrau & Seage, p. 22).