By assignment bias. For example, the participants in pen group may be older, smarter, Taller, or have higher socio-economic status the participants in another group one group may have a higher proportion of males or a bigger proportion of divorced individuals between groups that can compromise the research results