In an experiment on magazine advertisements, a researcher is examining visual displays to determine if one is more memorable. Only a small number of subjects will be avaliable for study. Should the researcher use a repeated measures design or independent samples design? and Why?