Volunteer Subjects as a Source of Sampling Bias

Many investigations require a sample of persons who are generally volunteer participants in the research. Since some refuse to participate, the question arises whether self-elimination of nonparticipants biases the sample obtained. This question was investigated in a study of engaged couples which secured data about participants and nonparticipants as to a number of characteristics. Participant were compared with participants and nonparticipant combined, the formed being the obtained sample and the latter the complete sample. The two samples differed on only one item, and this did not affect the findings.