Evaluation of the significance of student and institutional variables on the recruitment, retention, and graduation of minority students in senior institutions in higher education: A meta-analysis

There is no easy, succinct way to provide an objective, detailed description of the success of minorities in higher education. Qualitative and quantitative studies of hundreds of institutions and thousands of students, single-institution-subject research designs and case studies exist. Compared to whites, Native, Hispanic, and African Americans, score lower on achievement tests; enroll in colleges at a lower rate; and, attain fewer degrees. In light of the magnitude of the problem, the purpose of this research is to evaluate, through meta-analysis, the research that describes the success of underrepresented American minorities in senior institutions of higher education. Specifically, factors associated with the recruitment, admission, persistence, and graduation of African, Hispanic, and Native Americans at baccalaureate and advanced degree programs in the United States are identified and their significance is measured. Through meta-analysis, 113 studies reported between 1970 and 1997 were reviewed, and two-thirds provided information useful for synthesis utilizing the Stouffer method for averaging significance levels. The other third, consisting mostly of responses to surveys, was accomplished qualitatively. Briefly, institutional characteristics, family and background variables, and social and academic factors were found to be highly significant, many beyond the .0001 level, whereas finances and related concerns, work, financial aid, diversity, racism, and faculty variables were uniformly non-significant. Study design and subject characteristics did not contribute to the findings in a consistent, predictable fashion. An extensive catalog of data drawn from the primary studies is included in the Appendix. Major conclusions and recommendations include: (1) Incomplete reporting of data was a dominant obstacle. Basic descriptive statistics, validity and reliability indices and zero order correlation tables must be reported. (2) In a few instances, the outcome was listed as graduation, the paramount objective, whereas persistence to the second year was measured most often. (3) Definitions of variables that are culturally based and relevant for minorities were inconspicuous but may have practical and explanatory significance. This study lends credence to a holistic approach where a multitude of student academic and noncognitive characteristics are considered when deciding whom to admit and which supports to provide to maximize graduation rates