“…In this sense, the clusters may represent subgroupings on a continuum, and each subject's relative order within this continuum may be influenced by other unknown status variables. This research design controlled some possibilities noted in previous literature, particularly subject differences in SES (Ratusnik & Koenigsknecht, 1976), general exposure to SE (Bountress, 1983), and the geographic location of the subjects (Cole & Taylor, 1990;Washington & Craig, 1992a), but others await investigation.…”