Reasons for the lack of agreement among studies of persistence are considered, using a simplified causal model. Two major theoretical perspectives used in studying the educational careers of college entrants are summarized: (1) the Tinto or institutional persistence theory, which draws mainly on the discipline of education and applied institutional research; and (2) the discipline of sociology and social stratification. Attention is directed to differing conclusions about the importance of demographic and family background variables in explaining college persistence. The influence of how persistence is defined and the nature of the institution where data is collected is emphasized. A simplified model of selected background and post-college entry influences is estimated within four types of institutions (private liberal arts colleges, research/doctoral universities, public comprehensive four-year colleges, and two-year community/junior colleges. Persistence in a single college as well as persistence in higher education at several points in time are examined. Differences in results found in the analyses are tied to theoretical and statistical issues. Thirty-two references are cited. (Author/SW)