The Tension between Student
Persistence and Institutional Retention:
An Examination of the Relationship between First-
Semester GPA and Student Progression Rates of
First-Time Students (Session 529)
J. Hosch, Ph.D.
Braden J Hosch Ph D
Director of Institutional Research & Assessment
Central Connecticut State University, New Britain, CT
hoschbrj@ccsu.edu
Association for Institutional Research Annual Forum
Seattle, WA
May 26, 2008
This presentation is online at http://www.ccsu.edu/oira/research
Overview
First Semester
The Relationship Between First-Semester
GPA and Graduation and Retention Rates of
Full-Time, First-Time Students
(And two brief caveats)
The National Picture – Data from CSRDE
Case Study – Central Connecticut State U.
Conclusions and Implications
Major Findings
First semester full
First-semester grade point average of full-
time first-time students is:
very predictive of graduation and retention rates
this relationship is underreported
Implications
Students who perform well stay and graduate,
students who do not perform well tend to depart
Corollary: students get out of their education what
they put into it
Caveats
j
GPA is not just about Institutions should
academic not be let off the
performance, but also
encompasses factors hook for improving:
such as Instructional quality
Preparation Integration of co-
Eff t
Effort i l ti iti
curricular activities
Commitment Student contact with
j
Emotional adjustment faculty and staff
Social integration Ineffective policies
Financial stability and procedures
Etc.
Etc
CSRDE Study
Consortium for the Study of Retention Data
Exchange (CSRDE) collects institution-
reported data about progress and graduation
rates of full-time, first-time students
Data for cohort entering in 2000 published in
2007.
Institutions with incomplete data excluded
Institutions reporting <3% of entering cohort
earning first semester GPA<2.0 excluded
CSRDE Institutions Included in
Study Population
Control
Institutional Type Private (N) Public (N) Total (N) % of Total
Baccalaureate 15 20 35 11%
Master's 45 105 150 48%
Doctoral 11 112 123 39%
Other 4 3 7 2%
Total 75 240 315
Percent of Total 24% 76%
Institutional Success Rates by First
Semester GPA
100%
Mean Pct Retained or Graduated
d
90%
80%
70%
60%
50%
40% 82% 76%
65% 72% 65%
30% 53%
20% 45%
33%
10%
0%
3% to 10% 11% to 20% 21% to 30% g
31% and higher
(N=56) (N=155) (N=82) (N=22)
Proportion of Cohort Earning 1st Sem. GPA < 2.0
( )
One-Year Retention Rate (Mean) ( )
Six-Year Graduation Rate (Mean)
Error bars represent +/- one standard deviation
Source: 2006-07 CSRDE Retention Report
Scatterplot of Institutions and
Retention Rates
0 858 0.575 Below2.0
RetnRate = 0.858 - 0 575*Below2 0
R2 = 0.255
DF
N = 315 institutions, DF= 2, SSE = 1.76, SE = 0.075, p < 0.001; Excludes institutions reporting less than 3%
of the full-time, first-time cohort earned a first semester GPA below 2.0, on the basis that these institutions
are not representative of most post-secondary institutions. Data source (CSRDE, 2007)
Scatterplot of Institutions and
Graduation Rates
0 719 -1.14 Below2.0
GradRate = 0.719 -1 14*Below2 0
R2= 0.315
DF
N = 315 institutions, DF= 2, SSE = 5.16, SE = 0.128763, p < 0.001. Excludes institutions reporting less than
3% of the full-time, first-time cohort earned a first semester GPA below 2.0, on the basis that these
institutions are not representative of most post-secondary institutions. Data source (CSRDE, 2007)
Univariate Regression Models
Institutional One-Year Retention Rate
(Adj. R2=0.255) β S.E. t Sig.
Constant 0.858 0.011 80.04 ***
Percent of cohort with first semester GPA < 2.0 0.576
-0.576 0.055 10.41
-10.41 ***
Institutional Six-Year Graduation Rate
(Adj. R2=0.315) β S.E. t Sig.
Constant 0.719 0.018 39.22 ***
Percent of cohort with first semester GPA < 2.0 1.141
-1.141 0.095 12.05
-12.05 ***
Significant t 0 001
*** Si ifi t at p<0.001
Multivariate Regression Models
Institutional One-Year Retention Rate
(Adj. R2=0.367)
( ) β S
S.E. t Sig.
S
Constant 0.872 0.012 73.29 ***
Percent of cohort with first semester GPA < 2.0 -0.576 0.055 -10.41 ***
Baccalaureate institution 0 073
-0.073 0.014
0 014 5 36
-5.36 ***
Master’s institution -0.044 0.009 -5.10 ***
Public control 0.022 0.010 2.17 *
Institutional Six-Year Graduation Rate
(Adj. R2=0.380) β S.E. t Sig.
Constant 0.758 0.019 40.65 ***
Percent of cohort with first semester GPA < 2.0 -1.085 0.091 -11.97 ***
Master’s institution -0.079 0.015 -5.54 ***
Baccalaureate institution -0.097 0.023 -4.18 ***
institutional control was not significant for graduation rates in a stepwise regression
* Significant at p<0.05; *** Significant at p<0.001
Institutional Profile: Central
Connecticut State University
p y
Public – part of Connecticut State Univ. System
Carnegie 2005 Master’s-Larger Programs
New Britain, CT (Hartford MSA)
Fall 2007 Enrollment:
12,106 headcount (9,704 undergraduate, 23% residential);
full time
9,288 full-time equivalent enrollment
52% female; 16% minority
Full-time, first-time students: 1,469 (56% residential)
Full-time, new t
F ll ti f t d t
transfer students: 678
Six-year graduation rates:
,
44% full-time, first-time students
56% transfer students (full-time upon entry)
Institutional Progress Rates
One Year Six Year Full Time First
One-Year Retention Rate and Six-Year Graduation Rate of Full-Time First-
Time Students
90% National Peer
80%
79% Group Median 1-
80% 77% Year Retention
79% 72% Rate
78%
70% 75% National Peer
69% Group Median 6-
60% Year Graduation
Rate
46% 49%
50%
Institutional 1-
Year Retention
40% 44%
Rate
40%
40%
30%
Institutional 6-
Year Graduation
Rate
Fall of Entry
Six-Year Graduation Rates
Disaggregated (Entry F’99-F’01)
70%
63%
60%
Six-yea Graduation Rate
50% 48% 48% Institutional Avg.,
43% 41%
40% 35%
31%
30%
ar
20%
9%
10%
0%
Female Male Not Minority Minority 3.00-4.00 2.00-2.99 Below 2.0 or
WD
Gender Race/Ethnicity First Semester GPA
Six-Year Graduation Rate by First Semester
(Full Time First Time
GPA (Full-Time First-Time Students Entering
1999, 2000, 2001)
70%
63%
60%
-year Graduation Rate
50% 48%
R
Distribution of
40%
3.0-4.0 1st Semester
30% 2 0 2 99
2.0-2.99 GPAs
Below 2.0
20% 26% 32%
6-
9%
10%
42%
0%
1st Semester GPA
One-Year Retention Rates by First
Semester Grade Point Average
100%
90%
80%
70%
1999
60%
50% 2000
40% 2001
30% 2002
20% 2003
10% 2004
0% 2005
2006
First Semester GPA
Six-Year Graduation Rates by First
Semester Grade Point Average
80%
70%
60%
50%
40%
30% 1999
20% 2000
2001
10%
0%
First Semester GPA
Logistic Regression
Institutional One-Year Retention Rate Odds
(Cox & Snell R2=0.178, Nagelkerke R2=0.260) β S.E. Ratio Sig.
Constant -1.360 0.102
First semester GPA 1.082 0.044 2.952 ***
Prediction accuracy improves from 72% to 81%
Institution s Six-Year
Institution’s Six Year Graduation Rate Odds
Odd
(Cox & Snell R2=0.187, Nagelkerke R2=0.252) β S.E. Ratio Sig.
Constant -3.309 0.140
First semester GPA 1 194
1.194 0.051
0 051 3.301
3 301 ***
Prediction accuracy improves from 58% to 67%
Significant t <0 001
*** Si ifi t at p<0.001
Variables of HS rank, gender, and race/ethnicity were not observed to be significant
(p<0.05) in forward conditional entry.
Implications and Conclusions (1)
Institutions should focus on student success
and improved graduation rates will follow
Successful students stay and graduate,
unsuccessful students do not
Graduation and retention rates are indicators
of student success, not outcomes
Implications and Conclusions (2)
Focus on the first semester
FYE, early intervention programs
Student GPA over time remains relatively
consistent after controlling for attrition
Fall 2001 Fall 2002 Fall 2003 Fall 2004
Sem. Sem. Sem. Sem.
First Semester GPA N GPA N GPA N GPA N GPA
No GPA 33 -- 5 2.14 2 2.63 4 3.04
2.0
Below 2 0 258 1.12
1 12 103 1.82
1 82 58 2.11
2 11 41 2 53
2.53
2.00-2.49 255 2.26 213 2.33 178 2.45 155 2.67
2.50-2.99 345 2.73 284 2.59 256 2.65 236 2.82
3 00-3 49
3.00-3.49 240 3.21
3 21 204 2.89
2 89 184 2.99
2 99 177 3.11
3 11
3.50-4.00 141 3.71 114 3.33 101 3.34 95 3.45
Cohort Total 1272 2.50 923 2.60 779 2.74 708 2.93
Implications and Conclusions (3)
Consider student engagement, especially
time spent on academics
p y p p p g
Hrs per 7-day week FY students report preparing for class
(studying, reading, writing, doing homework or lab work,
analyzing data, rehearsing, and other academic activities),
50%
40%
30%
20% 45%
35%
10% 20%
0%
10 hours or less 11-20 hours per week 21 hours per week or
more
Source: NSSE National Results, 2007
Implications and Conclusions (4)
Grade inflation
Reassure faculty they are not being asked to
g y
grade differently, but instead to develop waysy
to prompt students to learn more effectively
Nevertheless, grade inflation is a possibility:
Undergraduate Grade Point Averages by Institution Type
50%
40% 17.5%
30%
15.5%
20% 11.2% A's and B's
10% 27.8%
10.9% 16.7% Mostly A's
0%
P bli 4
Public 4-year Private fit 4-year
P i t nonprofit 4 i t f fit
All private for-profit
Institution Type
Source: NCES (2007), National Postsecondary Student Aid Study 2003-04
Implications and Conclusions (5)
Reframe the national debate:
Balance notions of retention (institution’s
responsibility) with persistence (student’s
y) (
responsibility)
Student success (i.e. program completion)
requires a partnership among students,
institutions,
institutions and policymakers
The Tension between Student
Persistence and Institutional Retention:
An Examination of the Relationship between First-
Semester GPA and Student Progression Rates of
First-Time Students (Session 529)
J. Hosch, Ph.D.
Braden J Hosch Ph D
Director of Institutional Research & Assessment
Central Connecticut State University, New Britain, CT
hoschbrj@ccsu.edu
Association for Institutional Research Annual Forum
Seattle, WA
May 26, 2008
This presentation is online at http://www.ccsu.edu/oira/research