The author contends that the secondary analyst may have a different perspective and theoretical persuasion than the original investigator, may have the advantage of more current theory and practices than those originally available, may move beyond the substantive and methodological limitations of the original investigation, and may not be faced by the original time or funding constraints. Using existing data sets in the training of researchers also allows students to address problems relevant to their work without the need to devote enormous amounts of time to the collection of data. Secondary analysis may affect student interest in a field of research, and may expand opportunities to interact on topics of mutual interest. (Author/MV)