Institutional and Student Characteristics that Predict Graduation and Retention Rates Braden J. Hosch, Ph.D. Director of Institutional Research & Assessment November 4, 2008 North East Association for Institutional Research Annual Meeting Providence, RI This presentation and paper are online at http://www.ccsu.edu/oira
34
Embed
Institutional and Student Characteristics that Predict Graduation and Retention Rates
Institutional and Student Characteristics that Predict Graduation and Retention Rates. Braden J. Hosch, Ph.D. Director of Institutional Research & Assessment November 4, 2008 North East Association for Institutional Research Annual Meeting Providence, RI. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Institutional and Student Characteristics that Predict Graduation and Retention Rates
Braden J. Hosch, Ph.D.Director of Institutional Research & Assessment
November 4, 2008
North East Association for Institutional Research Annual MeetingProvidence, RI
This presentation and paper are online at http://www.ccsu.edu/oira
Graduation/retention rates of full-time, first-time students have serious limitations as metrics
Institutions participating in data sharing consortium have a special interest in progress rates
Institutional metrics include only students who enroll at these institutions
Institutional Profile: Central Connecticut State University Public – part of Connecticut State Univ. System Carnegie 2005 Master’s-Larger Programs New Britain, CT (Hartford MSA) Fall 2008 Enrollment:
Pct of Cohort Resided in Campus Housing .058 .016 3.656 ***
Addition of following factors can increase model power by 4.1% (R2=0.681): percent graduating in the top quartile of HS class; percent of cohort receiving student loans, and the percent of the cohort receiving federal grants; Percent of Cohort with 1st Term GPA Under 2.0.
Six-Year Graduation Rate Regression Model Using SAT ScoresInstitutional Six-Year Graduation Rate (Adj. R2=0.764) β S.E. t Sig.
(Constant) -60.401 6.670 -9.055
Combined Math and Verbal SAT score† .102 .006 17.125 ***
Pct of Cohort Resided in Campus Housing .161 .021 7.731 ***
Percent of Cohort w1st Term GPA Under 2.0 -.269 .071 -3.788 ***
Addition of following factors can increase model power by 4.5% (R2=0.811): Percent of all undergraduates who attend part-time, baccalaureate institution (dummy var.), percent graduating in the top quartile of HS class; percent of cohort receiving student loans, and the percent of the cohort receiving federal grants.
One-Year Retention Rate Regression Model NOT Using SAT ScoresInstitutional One-Year Retention Rate (Adj. R2=0.595) β S.E. t Sig.
(Constant) 55.50 1.363 40.72
Pct of Cohort Graduated in Top HS Quartile .286 .020 14.07 ***
Pct of Cohort Resided in Campus Housing .111 .016 6.99 ***
Addition of following factors can increase model power by 6.7% (R2=0.662): percent of the cohort receiving federal grants; expenditures on instruction and academic support per FTE; percent of cohort with a 1st term GPA under 2.0, public (dummy var.); percent of undergraduates who attend part-time, and percent of the cohort receiving student loans.
Six-Year Graduation Rate Regression Model NOT Using SAT ScoresInstitutional Six-Year Graduation Rate (Adj. R2=0.732) β S.E. t Sig.(Constant) 32.99 3.219 10.25
Pct of Cohort Graduated in Top HS Quartile .364 .031 11.83 ***
Pct of Cohort Resided in Campus Housing .211 .022 9.744 ***
Pct of Cohort that Received Federal Grants -.202 .039 -5.18 ***
Percent of Cohort w1st Term GPA Under 2.0 -.333 .075 -4.43 ***
Addition of following factors can increase model power by 4.5% (R2=0.811): Percent of all undergraduates who attend part-time, baccalaureate institution (dummy var.), percent graduating in the top quartile of HS class; percent of cohort receiving student loans, and the percent of the cohort receiving federal grants.
Six-Year Graduation Rate Regression Model Using Academic Inputs ONLYInstitutional Six-Year Graduation Rate (Adj. R2=0.790) β S.E. t Sig.(Constant) -45.13 6.935 -6.51
Mean Institutional SAT score .0786 .0080 9.79 ***
Pct of Cohort Resided in Campus Housing .158 .020 8.12 ***
Percent of Cohort 24+ years -1.04 .232 -4.50 ***
Pct of Cohort Graduated in Top HS Quartile .150 .038 3.90 ***
Implications and Conclusions (1)
Results confirm and extend previous research:
Most predictive factors: Admission inputs (SAT, followed by HS rank) Proportion living in campus housing First semester performance
Race, gender, and SES appear not to add significant predictive power AFTER controlling for above factors
Implications and Conclusions (2)
Policy implications: Evaluate institutional graduation rates in the
context of an expected graduation rate
Communicate realistic expectations to stakeholders
Implications and Conclusions (3)
Recognize the impact of academic inputs BEFORE and DURING college experience Selectivity is a significant factor that intersects
degree production as well as access; consider implications of resource allocation in context of degree yield rates
Set incentives to promote performance during college, e.g. loan forgiveness vs. merit-based scholarships
Implications and Conclusions (4)
Gaming the system - Institutions may continue to realize incentives to inflate grades
Public 4-year Private nonprofit 4-year All private for-profit0%5%
10%15%20%25%30%35%40%45%50%
10.9%16.7%
27.8%11.2%
15.5%
17.5%
Undergraduate Grade Point Averages by Institution Type
A's and B'sMostly A's
Institution Type
Implications and Conclusions (5)
Arms race in selectivity will be exposed by demographic change in next decade; downward pressure on graduation rates is likely
SOURCE: Knocking at the College Door (2008, Western Interstate Commission for Higher Education)Reproduced in The Chronicle of Higher Education 54.29.
Projections of Graduates of Public High Schools, by Racial and Ethnic Group in North East
White, Non-Hispanic
HispanicBlack, Non-HispanicAsian/Pacific Islander
Institutional and Student Characteristics that Predict Graduation and Retention Rates
Braden J. Hosch, Ph.D.Director of Institutional Research & Assessment
November 4, 2008
North East Association for Institutional Research Annual MeetingProvidence, RI
This presentation and paper are online at http://www.ccsu.edu/oira