This study of degree completion followed a national cohort of students from 10th grade (in 1980) through 1993 (High School and Beyond/Sophomore data base). Data included high school and college transcripts, test scores, and surveys. Using linear regression techniques, the study identified a model that accounts for about 43 percent of the variance in degree completion. The two most important variables identified were, first, academic resources a composite measure of the academic content and performance the student brings from secondary school, and, second, continuous enrollment. Major conclusions include the following: (1) since many students attend multiple institutions, institutional graduation rates are not very meaningful; (2) college admissions formulas that emphasize test scores and high school grade point average (rather than academic intensity and curriculum quality) are likely to produce lower degree completion rates; and (3) type and amount of remediation matters in relation to degree completion. Part 1 of the report constructs an index of student academic resources; Part 2 analyzes and/or reconstructs the major variables; Part 3 examines new attendance patterns and their significance; and Part 4 builds statistical models to explain findings. Five appendices include technical notes, additional tables, and examples. (Contains 125 references.) (DB)