Four statistics students discuss the following design questions; who has the best answer about type of sample that should be taken? "Researchers want to survey teenagers to see if there is a link between how often they eat dinner with the family and how welladjusted they are. If the researchers suspect that gender may play an important role in this relationship, should they take a convenience sample, a systematic sample, a cluster sample, or a stratified sample?" Adam: Convenience sample sounds good. If they take it in a convenient way, they can use a larger sample size. That will include lots of both sexes. Brittany: They should stratify the population into males and females, then take random samples, keeping track of gender along with the other variable. Carlos: Isn't that a systematic sample? You systematically pick one male, then one female, then another male, and another female, and so on. Dominique: If you want enough of both sexes, I think you should sample a cluster of males and a cluster of females, making the clusters as large as possible.