[decadence]

A U.S. Government study shows that over 70% of all those who join a fraternity/sorority graduate, while under 50% of all non-fraternity/sorority persons graduate. (also vaguely recall it mentioned this was a Univ. of Missouri study; unable to verify so far)

It wasn't exactly a US Government study... Though there was one in the past...

The NPC/NIC Research Initiative study has been conducted regularly since 1996. The 1998 study found (reported by Esther Thorson, director for the CASR) that since nearly 25 percent of the university's undergraduates are members of a fraternity or sorority, they expected that if Greeks were equally likely as non-Greek students to drop out then they would observe 25 percent of dropouts to have been in a fraternity or sorority. But their findings showed that only 18 percent of the dropouts identified themselves as having been members of a fraternity or sorority. "This would mean about a 28 percent higher retention rate than expected by chance," Dr. Thorson said in a news release. http://cyberbuzz.gatech.edu/nique/is...mpuslife4.html
It would appear this is the origin of the statistic.
Among the study's findings are that alumni of fraternities or sororities appear to be more likely to donate money to their alma maters than do other graduates, as well as feeling better about the social aspects of their undergraduate experience.
"The research findings show that young men and women who take an active part in a men's or women's fraternity or sorority are more likely to stay in college and to support their alma maters as alumni," said NPC Chairman Lissa Bradford and NIC President Idris R. Traylor in a news release.

Further sources related to student retention ( http://www.gammasigmaalpha.org/ ):
• Group living (residence halls and Greek houses) has a positive influence on retention and graduation (Carney, 1980)
• Fraternity residents are less likely to drop out (Eckland, 1964)
• More Greek transfer students graduate than non-members (Gibson, 1995)
• Greeks have higher graduation rates than non-Greeks (Iffert, 1958)
• Greek persistence and graduation rates were higher (Scott, 1965)

Carney, M. (1980). Persistence and graduation rates of Greek, independent, commuter and residence hall students: A nine semester study. Office of Student Affairs Research, 1979-1980 Report No. 45. Norman: University of Oklahoma. 3 pp. Abstract -- The results of this study support the positive socialization function that on-campus living offers students, particularly those with Greek affiliation. The fact that Greek students have lower academic ability as measured by standardized entrance tests, than residence hall non-Greek students, but yet have higher retention and graduation rates further supports the value of group living. {Single Institution study prior to 1990} [Abstract from Center for the Study of the College Fraternity, Annotated Bibliography, Research Studies and Articles 1980-1985]

Eckland, Bruce K. (1964). A source of error in college attrition studies. Sociology of Education, 38(1), 60-72.
Abstract -- Questionnaires returned by 1,180 former University of Illinois students were utilized to study predictors of college graduation. Academic careers of students over a ten-year period were examined. The general finding was that factors associated with a student’s early attrition from college have little connection with his or her later academic career. Students who resided in a fraternity were much less likely than others to dropout of the university, however dropouts who did not reside in a fraternity were more likely to return to college and graduate. [KRH]

Gibson, I. Jr. (1995). The relationship between graduation and selected variables among community college transfer students at Mississippi State University. Doctoral dissertation, Mississippi State University. Dissertation Abstracts International, 56 (6), 2089.
Abstract -- The purpose of this study was to determine whether a relationship existed between graduation and selected variables among community college students who transferred to Mississippi State University (MSU) during the fall semester of 1989. The selected variables were American College Test (ACT) score, chronological age, grade point average (GPA) prior to transferring, ethnicity (white or nonwhite), gender, credit hours transferred, Greek membership and living in a residence hall after transferring. Tinto's theory of persistence was selected as the theoretical perspective for understanding issues of students persisting to graduation in this study. A random sample of 200 subjects was selected from a population of 712 community college students who transferred to MSU. Data were collected from university records. Conclusions included: (1) The majority of the students who transferred to MSU from community colleges during the fall of 1989 graduated. The profile of community college students who persisted to graduation based on total numbers was: (a) ACT scores of 18 or higher, (b) approximately 20 years of age, (c) 60 or more credit hours earned prior to transferring to MSU, (d) white, (e) male, (f) not a member of a Greek organization, and (g) did not live in a residence hall. (2) Proportionately, more transfer students who were members of Greek organizations than non-Greek graduated. (3) Credit hours, GPA, and chronological age were the only variables that had a statistically significant correlation with persisting to graduation. [Author/KRH]

Iffert, Robert E. (1958). Retention and withdrawal of college students. Department of Health Education, and Welfare, Office of Education, Bulletin series, No. 1. Washington, D.C.: U.S. Government Printing Office.
Abstract -- Investigated the extent and character of retention, transfer, and withdrawal of 8,000 undergraduate students from a representative group of post-secondary institutions. The study covered 147 institutions of higher education in 46 states and the District of Columbia. Members of social fraternities who graduated had slightly, but not significantly, lower grades than non-fraternity male graduates. Members of social sororities had significantly higher grades than non-sorority women graduates. Students who were members or pledges of fraternities or sororities had better persistence records and graduation rates in the institutions of first registration than did nonmembers. Similarly, institutions having local or national social fraternities and sororities also had lower withdrawal rates than institutions without Greek letter organizations. [Author/KRH]

Scott, W. A. (1965). Values and organizations: A study of fraternities and sororities. Chicago: Rand McNally.
Abstract -- Studied ten fraternities and sororities at the University of Colorado over a one-year period to investigate how personal values enter into organizational processes. A couple of findings associated with actual academic achievement and student attrition are scattered throughout this well-indexed volume. The limited findings included: (1) persistence and graduation rates are higher among fraternity and sorority members than among non-members; and (2) when academic aptitude scores are equated with first semester grades, Greeks may show a decrease in academic performance relative to independents in their subsequent school years. Other findings concentrate on the values of students and groups of students including academic achievement and intellectualism as values of Greek organizations. 290 pages, 64 references. [KRH]
Abstract -- The results of this study support the positive socialization function that on-campus living offers students, particularly those with Greek affiliation. The fact that Greek students have lower academic ability as measured by standardized entrance tests, than residence hall non-Greek students, but yet have higher retention and graduation rates further supports the value of group living. {Single Institution study prior to 1990} [Abstract from Center for the Study of the College Fraternity, Annotated Bibliography, Research Studies and Articles 1980-1985]