A Step Toward College Success

A Step Toward College Success:
Assessing Attainment among Indiana?s Twenty-first Century Scholars
By
Edward P. St. John, University of Michigan,
and Jacob P.K. Gross, Glenda D. Musoba and Anna S. Chung, Indiana University
Prepared for
Lumina Foundation for Education
August 2005
2
Acknowledgments
Lumina Foundation for Education provided financial support for this research. The
Indiana Commission for Higher Education provided access to the Student Information
System college enrollment database, the Indiana Career and Postsecondary Advancement
Center provided ninth-grade survey data on Indiana students, and the State Student
Assistance Commission of Indiana provided information on student financial aid awards.
The authors sincerely appreciate this support. The analyses and opinions expressed in the
paper are the authors? and do not represent official policies or positions of Lumina
Foundation or the other organizations that supported the project.
1
Table of contents
Foreword ............................................................................................................................1
Executive summary ...........................................................................................................4
Introduction .......................................................................................................................8
Indiana?s Twenty-first Century Scholars program .....................................................10
Impact of the Twenty-first Century Scholars program ..............................................15
The 1999 cohort ....................................................................................................15
Academic preparation ...........................................................................................16
College choice .......................................................................................................19
Status of academic success: the cohort .................................................................19
Status of academic success: low-income students ................................................23
Conclusions ......................................................................................................................25
Appendices
1. The 1999 cohort: database design .....................................................................27
2. Attainment analyses for the entire cohort .........................................................30
3. Attainment analyses for low-income students ..................................................45
References ........................................................................................................................58
Endnotes ...........................................................................................................................59
About the authors ...........................................................................................................61
Publication information ..................................................................................................62
2
Foreword
Indiana?s Twenty-first Century Scholars program, a national model for combining
pre-college access intervention with financial aid, has improved college access for lowincome
students. The program is part of the state?s comprehensive approach to
expanding and equalizing opportunities for postsecondary education and has helped
Indiana increase its college access rates remarkably in recent years. (In 1986, Indiana
ranked 40th among states in the percentage of its high school graduates who went directly
on to college. By 2002, the state ranked ninth.). In 2002, Lumina Foundation funded a
study of the Scholars program and published a report titled Meeting the Access
Challenge: Indiana?s Twenty-first Century Scholars Program. That report documented
the program?s progress in enhancing access to higher education for low-income students,
but it did not examine whether it affected students? persistence or success in college.
Although the Scholars program was not designed to improve college persistence
rates, policy-makers and higher education leaders understand that access without
attainment is a hollow promise. Because our mission at Lumina Foundation is to enhance
access and success in postsecondary education, we are pleased to support this follow-up
study on the program. In this report, the principal authors of the original study continue
their research with the same cohort of students ? those who graduated from Indiana high
schools in 1999 ? to determine the impact of the Scholars program on college persistence
and attainment.
The findings show that the program has improved college persistence somewhat,
particularly with regard to students? attainment of two-year degrees. That is encouraging,
especially when one considers two important facts:
3| First, the Twenty-first Century Scholars program was designed and
implemented to encourage low-income students to enroll in postsecondary
education, not necessarily to help them persist or attain degrees. It?s
heartening to see any evidence, however slight, that the program may also
aid student persistence.| Second, the students this program was designed to assist (low-income
students eligible for the Federal Free and Reduced Lunch program) are
typically those who face the most ? and the most daunting ? barriers in
their pursuit of postsecondary education. For a variety of reasons, college
access and success rates among these students are persistently low, and
any effort that works to reverse that trend should be applauded.
Clearly, the Twenty-first Century Scholars program is no panacea for all of the
challenges that low-income students face. Much more must be done ? in Indiana and
throughout our nation ? to expand postsecondary access and success. Still, as the title of
this report suggests, the Scholars program helps low-income students take an important
early step toward college success. For that reason, we feel it is worthy of continued
support here in Indiana and increased attention and replication in other states.
Martha D. Lamkin
President and CEO,
Lumina Foundation for Education
4
Executive summary
The Twenty-first Century Scholars Program is an early-intervention program
intended to improve college enrollment among low-income students in Indiana. The
Scholars program provides support services ? plus a guarantee of grant aid equivalent to
public college tuition ? to students who pledge in eighth grade to meet five criteria:| Complete high school.| Maintain at least a C average.| Remain drug- and crime-free.| Apply for college and student financial aid.| Enroll in an Indiana college within two years of high school.
An earlier study of the Scholars program demonstrated its effectiveness in
enhancing access to higher education (St. John, Musoba, Simmons, & Chung, 2002).
That study, which focused on students in the 1999 cohort, showed that students who
participated in the program were more likely to enroll in college than were their
otherwise-average middle school peers. However, that 2002 study did not examine
whether the program had any effect on students? persistence or success in postsecondary
education. Although the Scholars program was not designed specifically to enhance
persistence or degree attainment, some positive effects could logically be expected, and
that is the aim of this current study: to identify, isolate and quantify those effects, if any.
This study again focuses on the 1999 cohort, examining the impact that a Scholars
award had on students? attainment status four years after enrollment in a public college
(i.e., at the end of the 2002-03 academic year). Our analyses of four-year attainment
status examined patterns among students in five categories: 1) those who were still
5
enrolled, 2) those who had attained two-year degrees, 3) those who had attained four-year
degrees, 3) those who had attained two- and four-year degrees, and 4) those who received
no degrees and 5) those no longer enrolled. The analyses controlled for the influence of
student background, high school contexts and preparation, the type of public college
attended, and college achievement. Students with Scholars awards and other types of aid
were compared to students who did not receive aid.
Findings from analyses of the entire cohort
Without statistical control, the analyses appeared to show that Scholars were less likely to
persist or attain degrees than were other students. Four years after high school, of the
students who took the Scholars pledge as eighth-graders, 50.1 percent had withdrawn
from college without earning a degree. The same was true of only 39.2 percent of
students who received other types of aid and just 35 percent of students who received no
aid. However, a more refined and accurate picture emerges when we control for the
influence of other variables. With such statistical control, the analyses showed that:
o Scholars were less likely to still be enrolled in college without a
degree than were students who received no aid. (On this measure,
there was no significant difference in the success rates of students
who received no aid and those who received types of aid other than
a Scholars award.)
o Among the three categories of students (Scholars, students with
other types of aid, and students with no aid), no significant
6
difference was found in their receipt of two-year degrees or their
receipt of both two- and four-year degrees.
o Students with other types of aid were more likely than students
who did not receive aid to have completed only a four-year degree,
but Scholars did not differ significantly on this measure from
students who received no student aid.
However, low-income students were less likely to still be enrolled and were less
likely to have attained degrees than students who did not apply for student aid. This
difference in attainment rates is no surprise when one considers the fact that low-income
students face challenges that their higher-income peers avoid. For this reason ? and
because the Scholars program was designed specifically to serve low-income students ? it
is important to compare the success rates of Scholars to those of low-income students
who did not participate in the program.
Findings from analyses of low-income students| After four years, 52.4 percent of eligible non-Scholars (low-income students
with other types of financial aid) had either left school or had failed to earn
even a two-year degree. This was true of only 47.1 percent of the Scholars.| Controlling for the influence of other variables, the analyses found that, after
four years and compared with their low-income peers who received other
types of financial aid:
o Scholars were more than twice as likely to have received two-year
degrees.
7
o Scholars did not differ significantly in their completion of four-year
degrees.
o Scholars did not differ significantly in persistence without a degree.| These findings indicate that, among low-income students, those who took the
Scholars pledge and enrolled in college had equal or better odds of academic
success than did their non-Scholar peers.
Conclusions
In combination with the earlier study of enrollment by the 1999 cohort, this study
indicates that Indiana?s Twenty-first Century Scholars Program improves the odds that
low-income students will enroll and succeed in college. Low-income students who take
the Scholars pledge as eighth-graders have demonstrably higher odds of enrolling in
college than do their peers. Scholars also are more likely to attain two-year degrees ? and
are no less likely to persist and attain four-year degrees ? than are other low-income
students who do not take the pledge.
Of course, much more needs to be done. Financial inequalities continue to affect
educational attainment among the residents of Indiana ? and every other state. Despite its
documented successes, the Scholars program can?t claim to have leveled the playing field
entirely. Still, it is a program that has proven worthy of emulation and deserving of
continued support.
8
Introduction
The State of Indiana has made remarkable improvements in college access since
the late 1980s. Several factors have contributed to this positive trend, including major
reforms in K?12 curricula (St. John, Musoba, & Chung, 2004), substantial investment in
student financial aid programs (St. John, Hu, & Weber, 2000, 2001), the development of
the Indiana Career and Postsecondary Advancement Center (ICPAC) [Hossler, Schmit, &
Vesper, 1999], and the creation of the Twenty-first Century Scholars program (St. John,
Musoba, Simmons, & Chung, 2002). In combination, these initiatives represent a
comprehensive approach to expanding and equalizing opportunities for college
preparation, access to higher education, and college persistence. This approach helped
Indiana make substantial gains in college access rates. (In 1986, the state ranked 40th in
the percentage of its high school graduates who went directly on to college; by 2002, it
ranked ninth.) Also, there is growing evidence that these reforms help equalize
opportunity for academic success in college in spite of rising college tuition (St. John,
Musoba, & Chung, 2004).
Although the Twenty-first Century Scholars program is not the sole reason for the
dramatic pattern of improvement, it is certainly part of the success story ? and it has
received national attention as a result. The National Governors Association has featured
the Scholars program at national meetings on best practices. The federal government has
pointed to Indiana?s example as it works to reform federal student financial aid programs
(Advisory Committee on Student Financial Assistance, 2003). A recent public television
program on college access, The College Track: America?s Sorting Machine, featured
9
Indiana?s reform strategies. And the state?s comprehensive approach to improving access
also was featured in a meeting organized by the Pathways to College Network.
While high school preparation and student financial aid both have been examined
for their effects on college persistence in Indiana (St. John, Musoba, & Chung, 2004), the
impact of the Twenty-first Century Scholars program on degree attainment had not been
examined before this study. After Spring 2003, it was possible to examine that impact
for the 1999 cohort of Scholars ? students who had previously been studied (St. John,
Musoba, Simmons, & Chung, 2002) and who, by Spring 2003, could have earned fouryear
degrees.
The attainment of two- or four-year degrees within four years of high school
graduation is an important indicator of academic success.1 To assess the impact of the
Scholars program, it is important not only to consider how the rates of Scholars?
attainment compare with those of other students, but also to measure (controlling for the
potential effects of other variables) whether, and to what extent, the support offered by
the program can be linked directly to degree attainment. This approach was used to study
the impact of the Scholars program on college access (enrollment) [St. John, Musoba,
Simmons, & Chung, 2002], and it can now be used to examine college success (degree
attainment).
Since other state programs can influence Twenty-first Century Scholars? success,
we must consider the state?s comprehensive approach to reform, not just the Scholars
program. Data limitations can complicate this type of comprehensive assessment,2 so it is
appropriate to consider the roles of the other Indiana programs and to incorporate
information from research on these programs. Since the attainment analyses included in
10
this report build on the earlier study of the 1999 cohort?s enrollment, this report follows
that cohort from high schools through degree attainment, summarizing the earlier study
and the current one. Statistical analyses on persistence are provided as supplementary
information (Appendices 2 and 3).
Indiana?s Twenty-first Century Scholars program
Initiated in 1990 by then-Gov. Evan Bayh, Indiana?s Twenty-first Century
Scholars program was the first state program to meet college tuition costs for middleschool
students who qualify for the Federal Free and Reduced Lunch program. To benefit
from this guarantee, students in the program pledge in the eighth grade to do all of the
following:| Graduate from an Indiana high school.| Maintain at least a 2.0 grade point average (GPA).| Apply for admission to an Indiana college.| Apply for financial aid.| Refrain from using illegal drugs and alcohol.| Refrain from committing a crime.| Enroll full time at an eligible Indiana college or university within two
years after high school.
In return for this pledge, the State of Indiana makes a commitment to:| Pay in full each Scholar?s tuition and fees (after other financial aid
awards) at any public college in Indiana or contribute a like portion of
tuition at an independent college.
11| Provide support services for Scholars, including tutoring, mentoring,
college visits and activities for parents.| Disseminate additional information on higher education and encourage
Scholars to pursue a college-preparatory curriculum.
The dual nature of the pledge process is designed to encourage low-income students
to take advantage of opportunities to prepare for college, as well as to enroll in college.
The Scholars program supplements the state grants Scholars receive as a consequence of
their aid eligibility, providing a small additional grant for full-need students and larger
supplemental awards for students with less need. Most students who received Scholars
awards also had full need, so their normal state grants were high, and the additional
award amounts were relatively modest. This pattern constrains the cost of the program for
the state and provides an added incentive to fund other state grant programs.
Figure 1 illustrates the role of the Twenty-first Century Scholars program within the
state?s framework for improving college access. There are six ways that Indiana?s
reforms can have a positive impact ? six links between these reforms and enhanced
educational opportunity:| The Twenty-first Century Scholars program guarantees adequate grant aid
to low-income eighth-graders who take a pledge to prepare for college
(Link 1). The report Meeting the Access Challenge: Indiana?s Twenty-first
Century Scholars Program (St. John, Musoba, Simmons, & Chung, 2002)
fully assesses the impact of the Scholars program (research summarized in
this report).
12| In addition to Indiana?s requirement that all high schools provide access to
the Core 40 curriculum (the college-preparatory curriculum in Indiana)
and honors diplomas, the state funds incentives to high schools to graduate
students with these diplomas (Link 2). Academic Preparation and College
Success (St. John, Musoba, & Chung, 2004) examines the impact of high
school diplomas on college choice and persistence. Scholars benefit from
this support, as do all students.3| The Indiana Career and Postsecondary Advancement Center (ICPAC,
renamed Learn More Resource Center) provided encouragement for
students to plan for postsecondary education, take preparatory curricula,
take college entrance exams, and apply for student aid and college
admission (Link 3). Before 1994, ICPAC also administered the services
for the Twenty-first Century Scholars program. The program is now
administered by the State Student Assistance Commission of Indiana
(SSACI).| The state need-based grant programs provide financial incentives for
students to complete a Core 40 or honors diploma (Link 4). The State of
Indiana recently took steps to make Core 40 the default diploma.| State grant programs provide support for enrollment and persistence
directly to students with financial need (Link 5). Meeting the Access
Challenge, summarized in this report, assesses the impact of the Scholars
program on college enrollment, including enrollment in four-year colleges.
13
The new analyses (Appendices 2 & 3) in this report provide the first
assessment of the impact of the Scholars program on college attainment.| State policy can affect the curricula that students actually complete, which,
in turn, can influence their college success. This means that state
education policies have a potential sustained, but indirect, effect on
college enrollment and persistence (Link 6). The indirect effects of state
curriculum reforms on college enrollment were considered in Academic
Preparation and College Success. To the extent that students gain access
to preparatory curricula, they can benefit from these reforms.5
The Twenty-first Century Scholars program is an integral part of Indiana?s
comprehensive approach for improving college access, especially for low-income
students. The information services provided by ICPAC were developed in tandem with
the Twenty-first Century Scholars program specifically to assist high-need students.
However, these services assist all students, and all students benefit from the
comprehensive approach that evolved in the state.
14
FIGURE 1. THE INDIANA MODEL
A Balanced Approach to College Access and Success
ACADEMIC POLICIES SOCIAL FACTORS FINANCE & ENCOURAGEMENT
Scholars Pledge
(Guarantee) 1
3
3
Student
Expectations &
Plans
High School Policy| Incentive
Funding| Curriculum
Requirements
Academic
Preparation
Enrollment in 2-Yr. &
4-Yr. College
College
Costs
Taking
Examination
Application
& Admission
Persistence
to Degree 5
5
Student
Financial Aid
State Grant
Programs
Family
Income &
Education
Source: Adapted Balanced Access Model, St. John, 2003
ICPAC
Direct link
Indirect link
3
4
2
6
Encouragement
3
6
15
Impact of the Twenty-first Century Scholars program
Tracking one cohort of students is one way to better understand the role that the
Twenty-first Century Scholars program plays in promoting academic success. The
analyses summarized below focus on the pipeline from ninth grade to college.
The 1999 cohort
Students in the 1999 cohort of Indiana high school students had the opportunity to
respond to a survey as high school freshmen. Most students responded to the ninth-grade
survey, as it was handed out and collected in their classrooms.6 The prior year, lowincome
students (those eligible for the Federal Free and Reduced Price Lunch program)
had been given the opportunity to sign up for the Twenty-first Century Scholars program.
To build a database for the 1999 cohort, we combined results from the ninth-grade grade
survey with records of students who had signed up as Scholars as eighth-graders. We also
added the following data:| Information for all students who applied for student financial aid as high
school seniors. (These data indicated that students had completed the
preparation process and had applied to private and/or out-of-state
colleges).| Information on curriculum at the high schools they attended (which
indicated the types of preparatory courses available to each student).| College enrollment information for 1999-2000 (to establish who had
enrolled in public colleges).
16| College enrollment information for 2000-01, 2001-02, and 2002-03 (to
indicate fourth-year enrollment and degree attainment).
Using this combination of data sources, it was possible to trace the educational progress
of students in the class. For example, we knew where students had enrolled in college
after high school and whether, after four years, students who had enrolled in state twoyear
and four-year colleges were still enrolled or had attained a degree. Below, we
examine evidence of preparation, college enrollment and choice, and status of academic
success after four years of college. In addition, we examine patterns of academic success
for low-income students who received aid, comparing Scholars? performance with that of
other students who received aid.
Academic preparation
Although the state of Indiana is taking steps to ensure that a college-preparatory
curriculum is the standard education offered all high school students, for the class of
1999 there was some variability in students? access to advanced curricula . Since the
attainment analyses are limited to students who enrolled initially in public colleges, we
considered the high school context for all students in the 1999 cohort (see Table 1),
broken down by whether students were Scholars, whether they received forms of aid
other than the Twenty-first Century Scholars program scholarship, and whether they did
not receive aid. These comparisons reveal the following:| A higher percentage of Scholars than non-Scholars attended urban high
schools.
17| Scholars graduated from high schools that had lower percentages of
students with honors diplomas, indicating lower levels of access to
advanced curricula.| A higher percentage of Scholars than non-Scholars attended high schools
with high percentages of students in the Federal Free and Reduced Price
Lunch program.| A higher percentage of Scholars than non-Scholars attended high schools
with high percentages of minority students.
Table 1. Comparison of Indicators of Preparation for High Schools Attended by All
Students, Scholars, Students Who Received Other Types of Student Aid and
Students Who Did Not Receive Student Aid (Based on the 1999 Cohort Who
Enrolled In Public Colleges)
Percentages of
students from ?
All
Students
Twenty-first
Century
Scholars
Students with
other student
aid
Students who
did not receive
aid
Urban high schools 14.2% 27.7% 15.4% 12.2%
Rural high schools 18.1% 15.1% 20.3% 16.3%
High schools with a
high percentage
(>26%) of honors
degrees
27.7% 18.5% 26.1% 29.7%
High schools with a
high percentage
(>17.7%) of Free and
Reduced Lunch
recipients
19.4% 42.6% 21.9% 15.7%
High schools with a
high percentage
(>13.7%) of minority
students
22.3% 37.9% 21.5% 22.1%
Although Indiana?s Twenty-first Century Scholars program does not yet reach all
eligible students, it does reach students who are attending schools that put them at risk of
not preparing for and enrolling in college. Many of the students in the 1999 cohort who
signed up as Scholars might have had lower odds of going to college had they not been
18
guaranteed student aid and, perhaps more important, had they not been provided
additional encouragement and academic support.
Earlier analyses of the 1999 cohort have examined the probability of applying for
student financial aid (see Table 2). Students who took the Scholars pledge were
consistently more likely to have applied for aid, indicating that the program has a positive
influence on pre-college awareness. Specifically, the otherwise-average Scholars
(controlling for other factors) were about 1.4 times more likely to apply for aid, with a
slightly larger odds ratio in the highest-poverty schools (St. John, Musoba, Simmons,
Chung, Schmit, & Peng, 2004). In addition, African-Americans who had taken the pledge
had higher odds of applying for aid (1.54 odds ratio) than did the otherwise average
African-American student. Further, white students in the program also had higher odds of
applying.
Table 2. Probabilities of Applying for Student Financial Aid, for Scholars
Compared to Others, for Students in the 1999 Cohort and Subgroups, Controlling
for Family Background, School Characteristics, and Student Achievement
Population Odds ratio Significance
Entire cohort 1.40 ***
Highest-poverty Schools 1.43 ***
Whites 1.28 ***
African-Americans 1.54 ***
*** p<.001, ** p<.01, * p<.05
Sources: St. John, Musoba, Simmons, Chung, Schmit, & Peng (2004); and Musoba (2004).
In combination, these findings indicate that the Scholars program helped students
overcome the risks associated with being in poverty and attending schools with large
proportions of at-risk students. These findings illustrate a general pattern of positive
program effects.
19
College choice
Earlier analyses of the impact of the Twenty-first Century Scholars program on
college choice by students in the 1999 cohort indicated that Scholars who took the steps
to prepare were more likely to enroll in colleges of all types (St. John, Musoba, Simmons,
& Chung, 2002). That research demonstrated, controlling for other variables in each case,
that:| Scholars were 4.43 times more likely to enroll in public four-year colleges
in 1999-2000 than non-Scholars.| Scholars were 6.37 times more likely to enroll in public two-year colleges
than were non-Scholars.| Scholars were 6.37 times more likely to enroll in Indiana?s private
colleges than were non-Scholars.| Scholars were 2.55 times more likely to be probable enrollees7 in out-ofstate
colleges.
These analyses provided evidence of a relationship between being a Scholar and
enrolling in college. In addition, supplemental analyses indicated more substantial
enrollment effects for African-American students than for white students (Musoba,
2004).
Status of academic success: the cohort
The new analyses in this section examine the impact of the Scholars program
scholarships on the postsecondary persistence and attainment of students in the 1999
cohort. The analyses were limited to students who enrolled initially in public colleges
20
because state-level data on private college students were not available. Approximately 10
percent (346) of Scholars enrolled in in-state private colleges, and 3 percent (118) of
Scholars expressed intentions to enroll in out-of-state institutions. Though these numbers
are small when compared with the whole Scholars sample, this lack of information might
have introduced a sample selection problem.8 Persistence and attainment by the entire
cohort (Appendix 2) are summarized below, followed by analyses for low-income
students.
The first step was to examine the persistence rate for students in the 1999 cohort
(see Table 3). Persistence was defined as having attained a college degree or being
enrolled in college at the end of the fourth year. After four years, 62.6 percent of the
students in the 1999 cohort who entered public colleges could be considered persisters by
this definition. This overall rate was higher than the rate for students with other forms of
aid (60.8 percent), higher than the rate for Scholars (49.9 percent), but lower than that for
students who did not receive aid (65.0 percent).
Table 3. Persistence Status in 2002-03 by Students in the 1999 Cohort
Who Enrolled Initially in Public Colleges
Persistence status All Scholars Other aid No aid
Nonpersisters 37.4% 50.1% 39.2% 35.0%
Persisters 62.6% 49.9% 60.8% 65.0%
Logistic regression analyses were used to examine the impact of the Twenty-first
Century Scholars program on persistence, controlling for other variables. These analyses
indicated that Scholars were less likely to be persisters after four years than were students
who did not receive aid, controlling for other variables (Table 2.2 in Appendix 2).
Students who received other forms of aid did not differ significantly in persistence from
students who did not receive student aid.
21
It is possible, however, that this model was not able to capture various
unobservable (and, we believe, income-related) characteristics that influence students?
persistence ? characteristics such as cultural capital, family support and availability of
information. To explore this issue further, we undertook two additional analyses: First,
we looked at our outcome differently by considering withdrawal against continuous
enrollment and degree attainment separately, rather than combined in a multinomial
logistic regression. Second, we performed the multinomial analysis on a fixed-income
student group: low-income students.
Before performing a multinomial logistic regression, we examined the association
between being a Scholar and educational attainment after four years (Table 4).
Descriptive statistics indicated that a modest percentage of Scholars (9.7 percent) had
obtained four-year degrees, very few (0.4 percent) had earned both two- and four-year
degrees, 4.2 percent had received two-year degrees, 35.7 percent had no degree yet still
persisted, and 50.1 percent of Scholars did not persist.
Table 4. Attainment Status in 2002-03by Students in the 1999 Cohort
Attainment status All Scholars* Other Aid No Aid
Nonpersister, no degree 37.4% 50.1% 39.2% 35.0%
Persister, no degree 42.0% 35.7% 40.6% 43.7%
Two-year degree 4.2% 4.2% 3.9% 4.5%
Both two- and four-year degrees 0.7% 0.4% 0.5% 0.9%
Four-year degree 15.7% 9.7% 15.8% 15.9%
*Total exceeds 100% due to rounding
We conducted a multinomial logistic regression to examine the impact of the Twentyfirst
Century Scholars program on attainment status (Table 2.4 in Appendix 2).
Consistent with the analysis of persistence, Scholars had lower odds of current
enrollment, but this was only a modest statistical association; students with other forms of
aid did not differ significantly from students who did not receive student aid. However,
22
Scholars did not differ significantly in their rate of degree attainment from students who
did not receive aid. Considering the unmeasured challenges that Scholars face ?
challenges that higher-income students typically avoid ?the fact that there is no
significant difference between Scholars and non-aid recipients in four-year degree
attainment is an important finding. Yet, Scholars had slightly lower odds of still being
enrolled after four years if they had not received a degree.9 This led us to the third
analyses ? one in which we restricted the sample to low-income students (Table 5). We
did this because we knew income and Scholar status were related.
Table 5. Summary of Multinomial Logistic Regression for Attainment
Status in 2002-03 by Students in the 1999 Cohort: Odds Ratios for
Aid Categories Compared to No Aid, Controlling for Other Variables
Scholars Other aid Attainment status
Odds ratio Sig. Odds ratio Sig.
Persister, no degree 0.771 ** 0.933 NS
Two-year degree 1.096 NS 0.952 NS
Both two- and four-year degrees 1.133 NS 0.991 NS
Four-year degree 1.014 NS 1.141 *
****p<0.001, ***p<0.01, **p<0.05, *p<0.10
Source: See Table 2.4 A to D in Appendix 2.
Indeed, students from low-income and lower-middle-income families were less
likely to be currently enrolled, or to have received a degree of any type, than were
students who did not apply for aid. This illustrates that, by the early 2000s, there were
income-related disparities among traditional-age students in persistence and degree
attainment, at least in Indiana. It is possible that the substantial investment in need-based
aid, including the Twenty-first Century Scholars program, was not sufficient to overcome
the disparities attributable to income. Therefore it is appropriate to examine the
persistence and attainment of low-income students specifically.
23
Status of academic success: low-income students
Those students whose families earned less than $30,621 and who received
financial aid ? Twenty-first Century Scholar program aid or otherwise ? were examined
next. Low-income financial aid recipients who were Scholars persisted at higher rates
over the four years (52.9 percent) than did low-income students who received aid but
were not Scholars (47.6 percent) [(Table 6). Scholars in the lowest-income category
persisted at higher rates (52.9 percent) than the average rate of persistence for all
Scholars (49.9 percent) (Table 3).
Table 6. Persistence Status in 2002-03 by Students in the
1999 Cohort Who Enrolled Initially in Public Colleges and
Were Low-income and Received Financial Ad
Persistence Status Scholars Other Aid
Nonpersisters 47.10% 52.40%
Persisters 52.90% 47.60%
An examination of educational attainment status shows that Scholars attained
four-year degrees, earned two-year degrees, and generally persisted at higher rates than
did non-Scholar aid recipients (Table 7). Low-income students who received aid but were
not Scholars obtained both two- and four-year degrees at a somewhat higher rate than
Scholars.
Table 7. Attainment Status in 2002-03by Low-income
Students in the 1999 Cohort Who Received Financial Aid
Attainment Status Scholars* Other Aid**
Nonpersister, no degree 47.10% 52.40%
Persister, no degree 37.90% 35.50%
Two-year degree 5.70% 3.70%
Both two- and four-year degrees 0.20% 0.40%
Four-year degree 9.00% 8.10%
*Due to rounding equals 99.9%
**Due to rounding equals 100.1%
24
A multinomial logistic regression for the lowest-income aid recipients was
conducted to explore educational attainment for Scholars and non-Scholars (Appendix
3).10 Results indicate that, controlling for other variables, Scholars do not differ
significantly from non-Scholar, lowest-income aid recipients in their likelihood to persist
and attain a four-year degree, including those with both two- and four-year degrees.
Perhaps most significant and consistent with the preceding descriptive analysis of
persistence, Scholars are more than twice as likely as non-Scholar aid recipients to attain
a two-year degree (Table 8).
Table 8. Summary of Multinomial Logistic Regression for
Attainment Status in 2002-03 by Students in the 1999 Cohort: Odds
Ratios for Twenty-first Century Scholars Compared to Aid Recipients
Who Were Not Scholars, Controlling for Other Variables
Scholars Attainment Status Odds ratio Sig.
Persister, no degree 0.946 NS
Two-year degree 2.026 **
Four-year degree and both two- and fouryear
degrees
0.998 NS
****p<0.001, ***p<0.01, **p<0.05, *p<0.10
Source: See Table 3.4 Parts A to C in Appendix 3.
25
Conclusions
The Twenty-first Century Scholars program is a comprehensive, earlyintervention
program intended to improve college enrollment among low-income
students in Indiana. An earlier study demonstrated that the program is effective in
enhancing access to higher education (St. John, Musoba, Simmons, & Chung, 2002).
That study, of students in the 1999 cohort, showed that students who took the Scholars
pledge were more likely to enroll in college than were other students. However, that
earlier study did not examine whether the program had any effect on students? persistence
or success in postsecondary education. Even though the Scholars program was not
designed specifically to enhance persistence or degree attainment, some positive effects
could logically be expected. This current study is an attempt to identify, isolate and
quantify such effects.
In this new study, we followed the 1999 cohort through the first four years of
college to assess the impact of the Twenty-first Century Scholars program on educational
attainment. The study had two major findings:
1. When the entire 1999 cohort was examined, it was evident that students in
the Twenty-first Century Scholars program persisted at a lower rate than
did students who received no student aid and students who received other
types of student aid. In addition, when the influence of other variables was
controlled for in multinomial logistic regression analyses, Scholars were
less likely to be currently enrolled than were students who received no aid.
In contrast, students who had received other forms of student aid were
more likely than non-Scholar aid recipients to have attained four-year
26
degrees within four years. Lower-middle-income students also were less
likely to have attained four-year degrees, and low-income students were
less likely to persist in general. Because Scholars were mostly from lowincome
families, it was important to consider persistence and degree
attainment specifically among low-income students.
2. When persistence by the population of low-income students was
examined, a different pattern was evident. Scholars had higher persistence
rates than did other low-income students who received financial aid. In
addition, controlling for the influence of other variables, Scholars had
higher odds of having attained two-year degrees than did low-income
students with other forms of aid (though they did not differ significantly in
their likelihood to be enrolled or to have attained four-year degrees).
Therefore, among low-income students, the very group that the Scholars
program was created to support, Scholars awards did have a modest
statistical association with academic success.
Clearly, financial inequalities continue to affect educational attainment among Indiana
residents. The Twenty-first Century Scholars program, however effective it has been,
can?t claim to have leveled the playing field entirely. However, low-income students who
take the Scholars pledge as eighth-graders do have demonstrably higher odds of enrolling
in college than do their peers. Scholars also are more likely to attain two-year degrees ?
and are no less likely to persist and attain four-year degrees ? than are other low-income
students who do not take the pledge.
27
Appendix 1
The 1999 cohort: database design
Six data sources contributed information for the merged database for the 1999
cohort used in the analyses on college enrollment in the earlier study, Meeting the Access
Challenge: Indiana?s Twenty-first Century program.
1. The Indiana Career and Postsecondary Advancement Center (ICPAC)
conducts an annual survey of all Indiana ninth-graders regarding their career
and educational aspirations, family living situation, current grades, etc. The
ninth-grade survey contained key predictor variables and students who would
fall into all of the possible outcome categories. With about an 80 percent
response rate, this provided a good representation of Indiana students as they
enter high school. This data set had 65,975 observations.
2. ICPAC also provided accurate records of which Indiana students had applied
for and enrolled in the Scholars program in eighth grade. This data set
contained 5,035 records of students who applied for the Scholars program.
3. The Indiana Commission for Higher Education collects from public colleges
and universities records regarding each student who attends an Indiana public
institution. The Student Information System (SIS) records include cumulative
grade point average, SAT scores, high school attended, degree plans, other
campuses attended, credits earned, campus living situation, family income,
demographic information, and financial aid records, including receipt of a
Twenty-first Century Scholars program scholarship. The SIS data set had
275,130 observations, including all undergraduates in the 1999-2000 database.
28
4. The State Student Assistance Commission of Indiana (SSACI), which
administers the state grants, provided access to students? Free Application for
Federal Student Aid (FAFSA) records, which offered data similar to the SIS
records for students who applied for admission to private and out-of-state
higher education institutions. For instance, the records indicated whether the
student received a state grant (indicating they went to an in-state institution) or
whether they applied for aid for an out-of-state school. SSACI data contained
records for all students from Indiana who applied for financial aid regardless
of attendance. The data set had 106,097 observations, including
undergraduates born in 1980-1982 (students who would be in ninth grade in
1995).
5. The Indiana Department of Education?s Web-based school data records
provided high school information such as the percentage of students receiving
Free or Reduced Price Lunches (a rough measure of poverty), the percentage
of graduates who earned an honors diploma, and other institutional variables.
6. The Indiana Commission for Higher Education (ICHE) Student Information
System (SIS) databases for 2000-01, 2001-02 and 2002-03 were integrated
into this database for the current study. The 2002-03 data set contained
362,469 observations and included data on cumulative grade point average,
SAT scores, high school attended, degree plans, other campuses attended,
credits earned, campus living situation, family income, demographic
information, and financial aid records, including receipt of a Twenty-first
29
Century Scholars program scholarship. The intervening years of data were
used in determining degrees earned.
The SIS and FAFSA records were merged using student identification numbers.
Because the ninth-grade surveys did not have identification numbers, a concatenation of
the student?s birth date, high school code, gender, ethnicity and home zip code was used
to match records. Progressive combinations of this concatenation from most restrictive
(i.e., all of the variables) to least restrictive were applied. In the first matches, variables
from the SIS data were matched. When that was exhausted, a concatenation of first and
last name, birth date, and gender was used to match remaining survey records with
FAFSA records. Following that, ninth-grade survey records were matched with the
Scholars data in a similar manner. A total of 21,615 observations were found in both the
SIS and FAFSA data files. All observations were retained in the file because students
who did not match with the SIS or FAFSA file have no evidence of higher education
enrollment and represented the reference group in the prior studies. Finally, high-schoollevel
variables were matched to the observations using the high school code unique to
each Indiana high school. This produced a database of 65,975 cases, of which 65,798
have no missing values and were used in the regression on college enrollments. Of this
data set, 25,856 students had attended a public two- or four-year college in Indiana in the
fall of 1999 and were considered in the persistence analysis. Of those, 24,062 cases had
no missing values and were included in the regressions on persistence.
30
Appendix 2
Attainment analyses for the entire cohort
31
Table 2.1 Descriptive Statistics for Variables in the Logistic Regression:
Indiana's Public Colleges 1999-2003
Variable Values
Number of
cases Percentage
Male and unknown 13,511 52.3% Gender
Female 12,341 47.7%
Missing or prefer not to answer 3,691 14.3%
Other minority 1,330 5.1%
African American 1,448 5.6%
Ethnicity
White 19,383 75.0%
Parent(s) went to college 16,187 62.6% Parental education level
No college for parent(s) or unknown parent
education
9,665 37.4%
Live with one parent 4,088 15.8%
Do not live with parent 516 2.0%
Who the student lived
with in ninth grade
Live with two parents or unknown living
situation
21,248 82.2%
Spanish or other language 207 0.8% Main language spoken in
the home English or unknown 25,645 99.2%
Missing/did not answer 1,285 5.0%
Mostly As 3,416 13.2%
Mixed Bs and Cs and Mostly Cs 6,806 26.3%
Mixed Cs and Ds or lower 1,462 5.7%
Grade point average in
ninth grade
Mixed As and Bs and Mostly Bs 12,883 49.8%
Low percentage of honors graduates 17,571 72.3% Percentage of graduates
in the student's high
school who earned an
honors diploma
High percentage of honors graduates 6,716 27.7%
Urban 3,675 14.2%
Rural 4,678 18.1%
Locale of high school
attended
Urban fringe, town and unknown 17,499 67.7%
Undecided/other/no response 3,579 13.8%
Aspire to a high school diploma or less 1,069 4.1%
Aspire to less than a two-year degree 871 3.4%
Aspire to a two-year degree 1,611 6.2%
Student?s aspirations in
ninth grade for education
completion
Aspire to a four-year degree or higher 18,722 72.4%
Relatively low minority concentration 19,443 77.7% Percentage in the
student's high school who
are minorities Relatively High minority concentration 5,593 22.3%
Low percentage of students 20,269 80.6% Percentage in the
student's school who
receive free or reducedprice
lunches
High percentage of students 4,883 19.4%
32
Low income (below $30,621) 3,544 13.7%
Lower-middle (between $30,621 and $52,719) 3,716 14.4%
Upper-middle (between $52,719 and $75,316) 3,774 14.6%
High income (above $75,316) 3,798 14.7%
Family income
quartiles
No reported income (did not apply for fin. aid) 11,020 42.6%
Self-supporting 423 1.6% Dependency status
Dependent on parents or indeterminate 25,429 98.4%
A 3,057 11.8%
B 10,896 42.1%
C 4,587 17.7%
Below C 6,562 25.4%
College cumulative
grade point average
Missing 750 2.9%
On campus 12,242 47.4% Living situation while
in college Live off campus, with parents or elsewhere 13,610 52.6%
Other four-year college 13,041 50.4%
Two-year college 4,019 15.5%
Institutional type
Research university 8,792 34.0%
21st Century Scholar with aid 855 3.3%
Other aid recipient 12,038 46.6%
Aid package
No aid or aid missing 12,959 50.1%
Nonpersisters 9,680 37.4% Persistence
Persisters 16,172 62.6%
33
Table 2.2 Logistic Regression Analysis of Persistence of Variables Associated With
Persistence: Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio Sig.
Female compared to male and unknown (zsex3) -0.013 0.03 0.987
When compared to Whites: (White)
Missing or prefer not to answer (Eth_Miss) -0.001 0.048 0.999
Other minorities (Min_oth) -0.023 0.076 0.977
African American (Black) -0.073 0.068 0.929
Parents went to college compared to no or unknown
parent higher education (zxfirst3) 0.181 0.033 1.199 ****
When compared to living with two parents or unknown
living situation (Live_no)
Live with one parent (Live_one) 0.034 0.163 1.035
Do not live with parent (Live_two) -0.048 0.042 0.953
Spanish or other language spoken at home compared
to English or unknown language spoken at home
(zxlangu3) -0.284 0.106 0.753 **
When compared to Mixed As and Bs and Mostly Bs
GPA in the 9th Grade (GPA_ABB)
Missing/did not answer (GPA_Miss) 0.128 0.086 1.137
Mostly As (GPA_A) 0.237 0.053 1.268 ****
Mixed Bs and Cs and Mostly Cs (GPA_BCC) -0.272 0.037 0.762 ****
Mixed Cs and Ds or lower (GPA_CD) -0.294 0.07 0.745 ****
When compared to families with no reported income
(did not apply for financial aid) (Inc_none)
Low income (below $30,621) (Inc_lo) -0.23 0.065 0.794 ****
Lower-middle (between $30,621 and $52,719) (Inc_lmid) -0.225 0.059 0.799 ****
Upper-middle (between $52,719 and $75,316) (Inc_umid) -0.112 0.056 0.894 **
High income (above $75,316) (Inc_high) -0.081 0.054 0.922
Self-supporting students when compared to students
dependent on parents or indeterminate dependency
status (depend3) -0.035 0.124 0.966
When compared to suburban/town/unknown:
(Loc_town)
Urban school locale (Loc_urb) 0.106 0.049 1.112 **
Rural school locale (Loc_rur) -0.046 0.04 0.955
Low percentage of students in the student's school
who receive free or reduced-price lunches (when
compared to high percentage) (zlunch2) 0.056 0.047 1.058
Low percentage minority students in high school
(compared to high percentage) (zminor2) -0.041 0.046 0.96
Low percentage honors diploma grads in high school
(when compared to high percentage) (zhonor2) -0.005 0.035 0.995
34
When compared to plans for a four-year degree or
higher: (Ed_4yr)
Undecided/other/no response (Ed_undec) -0.157 0.051 0.855 **
Aspire to a high school diploma or less (Ed_HS ) -0.233 0.077 0.792 **
Aspire to less than a two-year degree (Ed_lt2yr) -0.268 0.082 0.765 **
Aspire to a two-year degree (Ed_2yr) -0.157 0.061 0.855 **
When compared to students with a B cumulative
college GPA: (CGPA_B)
A (CGPA_A) 0.121 0.054 1.128 **
C (CGPA_C) -0.335 0.041 0.715 ****
Below C (CGPA_CLO) -1.351 0.037 0.259 ****
Missing (CGPA_MIS) -2.001 0.104 0.135 ****
Students who lived on campus compared to
students who lived off campus, with parents or
elsewhere (zhouse2) 0.371 0.035 1.449 ****
When compared to students who enrolled in a
public research university (En_res )
Other four-year college (En_4yr ) -0.258 0.04 0.773 ****
Two-year college (En_2yr) -1.294 0.053 0.274 ****
When compared to students who received no
financial aid (Aid_none)
21st Century Scholar with aid (Aid_21 ) -0.188 0.093 0.829 **
Other aid recipient (Aid_othe) 0.007 0.046 1.007
****p<0.001, ***p<0.01, **p<0.05, *p<0.10
N of cases= 24,062
-2 Log likelihood= 26,825.443
Chi-square= 5062.015**
Nagelkerke= 0.258
*1790 cases were excluded from the regression due to missing values.
35
Table 2.3 Descriptive Statistics for Variables in the Multinomial Regression:
Indiana's Public Colleges 1999-2003
Variable Values
Number of
cases Percentage
Male and unknown 13,511 52.3% Gender
Female 12,341 47.7%
Missing or prefer not to answer 3,691 14.3%
Other minority 1,330 5.1%
African American 1,448 5.6%
Ethnicity
White 19,383 75.0%
Parent(s) went to college 16,187 62.6% Parental education level
No college for parent(s)/unknown par. ed. 9,665 37.4%
Live with one parent 4,088 15.8%
Do not live with parent 516 2.0%
Who the student lived with
in ninth grade
Live with 2 parents/unknown living situation 21,248 82.2%
Spanish or other language 207 0.8% Main language spoken in
the home English or unknown 25,645 99.2%
Missing/did not answer 1,285 5.0%
Mostly As 3,416 13.2%
Mixed Bs and Cs and Mostly Cs 6,806 26.3%
Mixed Cs and Ds or lower 1,462 5.7%
Grade point average in
ninth grade
Mixed As and Bs and Mostly Bs 12,883 49.8%
Low percentage of honors graduates 17,571 72.3% Percentage of graduates in
student's high school who
earned honors diplomas High percentage of honors graduates 6,716 27.7%
Urban 3,675 14.2%
Rural 4,678 18.1%
Locale of high school
attended
Urban fringe, town and unknown 17,499 67.7%
Undecided/other/no response 3,579 13.8%
Aspire to a high school diploma or less 1,069 4.1%
Aspire to less than a two-year degree 871 3.4%
Aspire to a two-year degree 1,611 6.2%
Student's aspirations in
ninth grade for education
completion
Aspire to a four-year degree or higher 18,722 72.4%
Relatively low minority concentration 19,443 77.7% Percentage in the student's
high school who are
minorities Relatively High minority concentration 5,593 22.3%
Low percentage of students 20,269 80.6% Percentage in the student's
school who receive free or
reduced-price lunches High percentage of students 4,883 19.4%
Low income (below $30,621) 3,544 13.7%
Lower-middle (between $30,621 & 52,719) 3,716 14.4%
Upper-middle (between $52,719 & 75,316) 3,774 14.6%
High income (above $75,316) 3,798 14.7%
Family income quartiles
No reported income (did not apply for fin. aid) 11,020 42.6%
Self-supporting 423 1.6% Dependency status
Dependent on parents or indeterminate 25,429 98.4%
36
A 3,057 11.8%
B 10,896 42.1%
C 4,587 17.7%
Below C 6,562 25.4%
College cumulative grade
point average
Missing 750 2.9%
On campus 12,242 47.4% Living situation while in
college Live off campus, with parents or elsewhere 13,610 52.6%
Other four-year college 13,041 50.4%
Two-year college 4,019 15.5%
Institutional type
Research university 8,792 34.0%
21st Century Scholar with aid 855 3.3%
Other aid recipient 12,038 46.6%
Aid package
No aid 12,959 50.1%
Nonpersister, no degree 8,336 32.2%
Persister, no degree 10,855 42.0%
Two-year degree 1,874 7.2%
Both Two- and Four-year degrees 190 0.7%
Enrollment outcomes
Four-year degree 4,597 17.8%
37
Table 2.4, Part A. Multinomial Regression Analysis of Variables
Associated With Completion of Four-year Degree:
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio
Sig.
Female compared to male and unknown (zsex2) -0.19 0.047 0.827 ****
When compared to whites: (zxsethn2)
Missing or prefer not to answer -0.022 0.077 0.978
Other minorities -0.201 0.111 0.818 *
African American 0.039 0.128 1.04
Parents went to college compared to no or unknown
parent higher education (zxfirst2) 0.191 0.05 1.211 ****
When compared to living with two parents or
unknown living situation (zxslive2)
Live with one parent -0.158 0.069 0.853 **
Do not live with parent -0.167 0.179 0.846
Spanish or other language spoken at home compared
to English or unknown language spoken at home
(zxlangu2)
-0.507 0.303 0.602 *
When compared to Mixed As and Bs and Mostly Bs
GPA in the 9th Grade (zxgpa2)
Missing/did not answer 0.431 0.137 1.539 **
Mostly As 0.384 0.067 1.468 ****
Mixed Bs and Cs and Mostly Cs -0.227 0.066 0.797
Mixed Cs and Ds or lower 0.182 0.115 1.2
When compared to families with no reported income
(did not apply for financial aid) (zincome)
Low income (below $30,621) -0.17 0.111 0.844
Lower-middle income (between $30,621 and $52,719) -0.205 0.097 0.814 **
Upper-middle income (between $52,719 and $75,316) 0.022 0.088 1.022
High income (above $75,316) 0.108 0.079 1.114
Self-supporting students when compared to students
dependent on parents or indeterminate dependency
status (depend2)
-0.384 0.285 0.681
When compared to suburban/town/unknown:
(zlocale2)
Urban school locale 0.019 0.076 1.019
Rural school locale 0.027 0.061 1.027
Low percentage of students in the student's school
who receive free or reduced-price lunches (when
compared to high percentage) (zlunch)
0.057 0.075 1.059
Low percentage minority students in high school
(compared to high percentage) (zminor) 0.055 0.07 1.057
Low percentage honors diploma grads in high school
(when compared to high percentage) (zhonor) -0.087 0.051 0.917
38
When compared to plans for a four-year degree or
higher: (zedexp2)
Undecided/other/no response -0.209 0.085 0.812 **
Aspire to a high school diploma or less -0.122 0.13 0.885
Aspire to less than a two-year degree -0.111 0.14 0.895
Aspire to a two-year degree -0.188 0.112 0.829 *
When compared to students with a B cumulative
college GPA: (zcumgpa)
A 0.634 0.065 1.886 ****
C -1.175 0.069 0.309
Below C -3.748 0.14 0.024 ****
Missing -20.706 2227 0.000
Students who lived on campus compared to students
who lived off campus, with parents or elsewhere
(zhousing)
0.706 0.054 2.025 ****
When compared to students who enrolled in a public
research university (zenroll)
Other four-year college -0.921 0.056 0.398 ****
Two-year college -4.568 0.265 0.01
When compared to students who received no
financial aid (zaid2)
21st Century Scholar with aid 0.014 0.163 1.014
Other aid recipient 0.132 0.072 1.141 *
Dependent variable: enrollment outcome by 2003
****p<0.001, ***p<0.01, **p<0.05, **p<0.10
N of cases= 24,062?
-2 Log likelihood= 37,1916.855
Chi-square= 14.955**
McFadden= 0.160
Nagelkerke= 0.350
?1790 cases were excluded from the regression due to missing values.
39
Table 2.4, Part B. Multinomial Regression Analysis of Variables
Associated With Completion of Both Two- and Four-Year Degree:
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio Sig.
Female compared to male and unknown (zsex3) 0.607 0.17 1.835 ****
When compared to Whites: (zxsethn2)
Missing or prefer not to answer 0.13 0.26 1.138
Other minorities -0.295 0.432 0.744
African American 0.053 0.5 1.055
Parents went to college compared to no or unknown
parent higher education (zxfirst2) 0.212 0.18 1.236
When compared to living with two parents or
unknown living situation (zxslive2)
Live with one parent 0.157 0.236 1.17
Do not live with parent -1.172 1.017 0.31
Spanish or other language spoken at home compared
to English or unknown language spoken at home
(zxlangu2) -0.358 1.029 0.699
When compared to Mixed As and Bs and Mostly Bs
GPA in the 9th Grade (zxgpa2)
Missing/did not answer -0.01 0.475 0.99
Mostly As 0.364 0.221 1.439 *
Mixed Bs and Cs and Mostly Cs -0.185 0.238 0.831
Mixed Cs and Ds or lower -0.102 0.392 0.903
When compared to families with no reported income
(did not apply for financial aid) (zincome)
Low income (below $30,621) -0.676 0.426 0.509
Lower-middle income (between $30,621 and $52,719) -1.11 0.407 0.33 **
Upper-middle income (between $52,719 and $75,316) -0.655 0.34 0.519 *
High income (above $75,316) -0.568 0.3 0.567 *
Self-supporting students when compared to students
dependent on parents or indeterminate dependency
status (depend2) 0.628 0.789 1.873
When compared to suburban/town/unknown:
(zlocale2)
Urban school locale 0.087 0.282 1.091
Rural school locale 0.386 0.202 1.471 *
Low percentage of students in the student's school
who receive free or reduced-price lunches (when
compared to high percentage) (zlunch) 0.463 0.303 1.589
Low percentage minority students in high school
(compared to high percentage) (zminor) -0.171 0.247 0.843
Low percentage honors diploma grads in high school
(when compared to high percentage) (zhonor) 0.004 0.183 1.004
40
When compared to plans for a four-year degree or
higher: (zedexp2)
Undecided/other/no response 0.151 0.279 1.163
Aspire to a high school diploma or less 0.121 0.439 1.128
Aspire to less than a two-year degree 0.537 0.406 1.71
Aspire to a two-year degree 0.097 0.37 1.102
When compared to students with a B cumulative
college GPA: (zcumgpa)
A 1.036 0.184 2.817 ****
C -1.117 0.284 0.327 ****
Below C -3.563 0.59 0.028 ****
Missing -21.057 0 0.000
Students who lived on campus compared to students
who lived off campus, with parents or elsewhere
(zhousing) 0.234 0.194 1.263
When compared to students who enrolled in a public
research university (zenroll)
Other four-year college -0.792 0.208 0.453 ****
Two-year college -1.662 0.347 0.19 ****
When compared to students who received no
financial aid (zaid2)
21st Century Scholar with aid 0.125 0.662 1.133
Other aid recipient -0.009 0.267 0.991
Comparison group for enrollment outcome is nonpersisters
****p<0.001, ***p<0.01, **p<0.05, **p<0.10
N of cases= 24,062?
-2 Log likelihood= 37,1916.855
Chi-square= 14.955**
McFadden= 0.160
Nagelkerke= 0.350
?1790 cases were excluded from the regression due to missing values.
41
Table 2.4, Part C. Multinomial Regression Analysis of Variables
Associated With Completion of Two-Year Degree:
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio
Sig.
Female compared to male and unknown (zsex3) -0.063 0.069 0.939
When compared to Whites: (zxsethn2)
Missing or prefer not to answer -0.148 0.114 0.862
Other minorities 0.147 0.146 1.159
African American -0.194 0.203 0.824
Parents went to college compared to no or
unknown parent higher education (zxfirst2) -0.03 0.077 0.97
When compared to living with two parents or
unknown living situation (zxslive2)
Live with one parent -0.031 0.098 0.97
Do not live with parent -0.425 0.285 0.654
Spanish or other language spoken at home
compared to English or unknown language spoken
at home (zxlangu2) 0.477 0.315 1.612
When compared to Mixed As and Bs and Mostly Bs
GPA in the 9th Grade (zxgpa2)
Missing/did not answer -0.103 0.194 0.903
Mostly As 0.038 0.118 1.038
Mixed Bs and Cs and Mostly Cs -0.489 0.086 0.613 ****
Mixed Cs and Ds or lower -0.998 0.193 0.369 ****
When compared to families with no reported
income (did not apply for financial aid) (zincome)
Low income (below $30,621) -0.606 0.148 0.545 ****
Lower-middle income (between $30,621 and $52,719) -0.597 0.134 0.55 ****
Upper-middle income (between $52,719 and $75,316) -0.386 0.124 0.68 **
High income (above $75,316) -0.399 0.127 0.671 **
Self-supporting students when compared to
students dependent on parents or indeterminate
dependency status (depend2) 0.249 0.26 1.283
When compared to suburban/town/unknown:
(zlocale2)
Urban school locale 0.105 0.111 1.111
Rural school locale -0.021 0.086 0.979
Low percentage of students in the student's school
who receive free or reduced-price lunches (when
compared to high percentage) (zlunch) -0.059 0.106 0.942
Low percentage minority students in high school
(compared to high percentage) (zminor) 0.242 0.111 1.274 **
Low percentage honors diploma grads in high
school (when compared to high percentage)
(zhonor) 0.293 0.083 1.34 ****
42
When compared to plans for a four-year degree or
higher: (zedexp2)
Undecided/other/no response 0.012 0.113 1.012
Aspire to a high school diploma or less 0.046 0.176 1.047
Aspire to less than a two-year degree -0.122 0.194 0.885
Aspire to a two-year degree 0.219 0.123 1.245 *
When compared to students with a B cumulative
college GPA: (zcumgpa)
A -0.067 0.109 0.935
C -0.507 0.093 0.602 ****
Below C -1.654 0.098 0.191 ****
Missing -2.342 0.298 0.096 ****
Students who lived on campus compared to
students who lived off campus, with parents or
elsewhere (zhousing) -0.222 0.083 0.801 **
When compared to students who enrolled in a
public research university (zenroll)
Other four-year college -0.039 0.096 0.962
Two-year college 0.244 0.112 1.276 **
When compared to students who received no
financial aid (zaid2)
21st Century Scholar with aid 0.091 0.215 1.096
Other aid recipient -0.05 0.103 0.952
Comparison group for enrollment outcome is nonpersisters
****p<0.001, ***p<0.01, **p<0.05, **p<0.10
N of cases= 24,062?
-2 Log likelihood= 37,1916.855
Chi-square= 14.955**
McFadden= 0.160
Nagelkerke= 0.350
?1790 cases were excluded from the regression due to missing values.
43
Table 2.4, Part D. Multinomial Regression Analysis of Variables
Associated With Persisters Who Have Not Completed a Degree:
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio Sig.
Female compared to male and unknown (zsex3) 0.033 0.032 1.033
When compared to Whites: (zxsethn2)
Missing or prefer not to answer 0.018 0.05 1.018
Other minorities -0.079 0.072 0.924
African American -0.025 0.079 0.975
Parents went to college compared to no or
unknown parent higher education (zxfirst2) 0.197 0.035 1.218 ****
When compared to living with two parents or
unknown living situation (zxslive2)
Live with one parent -0.031 0.044 0.969
Do not live with parent -0.281 0.113 0.755 **
Spanish or other language spoken at home
compared to English or unknown language spoken
at home (zxlangu2) 0.07 0.169 1.073
When compared to Mixed As and Bs and Mostly Bs
GPA in the 9th Grade (zxgpa2)
Missing/did not answer 0.083 0.091 1.086
Mostly As 0.172 0.056 1.188 **
Mixed Bs and Cs and Mostly Cs -0.269 0.039 0.764 ****
Mixed Cs and Ds or lower -0.339 0.075 0.712 ****
When compared to families with no reported
income (did not apply for financial aid) (zincome)
Low income (below $30,621) -0.202 0.069 0.817 **
Lower-middle income (between $30,621 and $52,719) -0.182 0.062 0.834 **
Upper-middle income (between $52,719 and $75,316) -0.109 0.059 0.897 *
High income (above $75,316) -0.111 0.057 0.895 *
Self-supporting students when compared to
students dependent on parents or indeterminate
dependency status (depend2) -0.046 0.132 0.955
When compared to suburban/town/unknown:
(zlocale2)
Urban school locale 0.13 0.051 1.138 **
Rural school locale -0.065 0.042 0.937
Low percentage of students in the student's school
who receive free or reduced-price lunches (when
compared to high percentage) (zlunch) 0.061 0.049 1.063
Low percentage minority students in high school
(compared to high percentage) (zminor) -0.079 0.048 0.924
Low percentage honors diploma grads in high
school (when compared to high percentage)
(zhonor) 0.001 0.037 1.001
44
When compared to plans for a four-year degree or
higher: (zedexp2)
Undecided/other/no response -0.17 0.053 0.844 **
Aspire to a high school diploma or less -0.3 0.083 0.741 ****
Aspire to less than a two-year degree -0.333 0.089 0.716 ****
Aspire to a two-year degree -0.218 0.066 0.804 **
When compared to students with a B cumulative
college GPA: (zcumgpa)
A -0.198 0.06 0.821 **
C -0.101 0.043 0.904 **
Below C -1.018 0.039 0.361 ****
Missing -1.689 0.108 0.185 ****
Students who lived on campus compared to
students who lived off campus, with parents or
elsewhere (zhousing) 0.358 0.036 1.43 ****
When compared to students who enrolled in a
public research university (zenroll)
Other four-year college -0.058 0.042 0.944
Two-year college -1.137 0.057 0.321 ****
When compared to students who received no
financial aid (zaid2)
21st Century Scholar with aid -0.26 0.099 0.771 **
Other aid recipient -0.007 0.049 0.993
Comparison group for enrollment outcome is nonpersisters
****p<0.001, ***p<0.01, **p<0.05, **p<0.10
N of cases= 24,062?
-2 Log likelihood= 37,1916.855
Chi-square= 14.955**
McFadden= 0.160
Nagelkerke= 0.350
?1790 cases were excluded from the regression due to missing values.
45
Appendix 3
Attainment analyses for low-income students
46
Table 3.1. Descriptive Statistics for Variables in the Logistic Regression for
Enrollment Outcomes for Low-income, Financial Aid Recipients:
Indiana's Public Colleges 1999-2003
Variable Values
Number of
cases Percentage
Male and unknown 1,338 42.2% Gender
Female 1,829 57.8%
Missing or prefer not to answer 535 16.9%
Other minority 201 6.3%
African American 435 13.7%
Ethnicity
White 1,996 63.0%
Parent(s) went to college 2,461 77.7% Parental education level
No college for parent(s) or
unknown parent education
706 22.3%
Live with one parent 988 31.2%
Do not live with parent 152 4.8%
Who the student lived
with in ninth grade
Live with two parents or unknown
living situation
2,027 64.0%
Spanish or other language 26 0.8% Main language spoken in
the home English or unknown 3,141 99.2%
Missing/did not answer 206 6.5%
Mostly As 263 8.3%
Mixed Bs and Cs and Mostly Cs 1,081 34.1%
Mixed Cs and Ds or lower 244 7.7%
Grade point average in
ninth grade
Mixed As and Bs and Mostly Bs 1,373 43.4%
Low percentage of honors
graduates
2,370 79.6% Percentage of graduates
in the student's high
school who earned an
Honors diploma High percentage of honors
graduates
609 20.4%
Urban 650 20.5%
Rural 617 19.5%
Locale of high school
attended
Urban fringe, town and unknown 1,900 60.0%
Undecided/other/no response 540 17.1%
Aspire to a high school diploma or
less
177 5.6%
Aspire to less than a two-year
degree
148 4.7%
Aspire to a two-year degree 271 8.6%
Student's aspirations in
ninth grade for education
completion
Aspire to a four-year degree or
higher
2,031 64.1%
Relatively low minority
concentration
2,115 70.7% Percentage of students
in the student's high
school who are
minorities
Relatively High minority
concentration
878 29.3%
47
Low percentage of
students
2,015 66.5% Percentage of students in the student's
school who receive free or reduced-price
lunches
High percentage of
students
1,015 33.5%
Self-supporting 355 11.2% Dependency status
Dependent on parents
or indeterminate
2,812 88.8%
A 225 7.1%
B 1,113 35.1%
C 557 17.6%
Below C 1,111 35.1%
College cumulative grade point average
Missing 161 5.1%
On campus 1,212 38.3% Living situation while in college
Live off campus, with
parents or elsewhere
1,955 61.7%
Other four-year
college
1,631 51.5%
Two-year college 1,114 35.2%
Institutional type
Research university 422 13.3%
21st Century Scholar
with aid
391 12.3% Aid package
Other aid recipient 2,776 87.7%
Nonpersisters 1,623 51.2% Persistence
Persisters 1,544 48.8%
48
Table 3.2 Logistic Regression Analysis of Variables Associated with Persistence by
Low-income, Financial Aid Recipients Enrolled in
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio Sig.
Female compared to male and unknown (zsex3) -0.063 0.087 0.939
When compared to Whites: (white)
Missing or prefer not to answer (Eth_Miss) 0.52 0.135 1.682 ****
Other minorities (Min_oth) 0.265 0.16 1.303 *
African American (Black) -0.002 0.177 0.998
Parents went to college compared to no or
unknown parent higher education (zxfirst3) 0.107 0.104 1.113
When compared to living with two parents or
unknown living situation (Live_no)
Live with one parent (Live_one) 0.019 0.481 1.019
Do not live with parent (Live_two) 0.223 0.095 1.25 **
Spanish or other language spoken at home
compared to English or unknown language spoken
at home (zxlangu3) -0.241 0.211 0.786
When compared to Mixed As and Bs and Mostly Bs
GPA in the 9th Grade (GPA_ABB)
Missing/did not answer (GPA_Miss) -0.585 0.228 0.557 **
Mostly As (GPA_A) 0.232 0.165 1.261
Mixed Bs and Cs and Mostly Cs (GPA_BCC) -0.152 0.102 0.859
Mixed Cs and Ds or lower (GPA_CD) -0.338 0.186 0.713 *
Self-supporting students when compared to
students dependent on parents or indeterminate
dependency status (depend3) -0.009 0.138 0.991
When compared to suburban/town/unknown:
(Loc_town)
Urban school locale (Loc_urb) 0.451 0.125 1.57 ****
Rural school locale (Loc_rur) -0.1 0.111 0.905
Low percentage of students in the student's school
who receive free or reduced-price lunches (when
compared to high percentage) (zlunch2) 0.269 0.119 1.308 **
Low percentage minority students in high school
(compared to high percentage) (zminor2) 0.013 0.14 1.013
Low percentage honors diploma grads in high
school (when compared to high percentage)
(zhonor2) -0.321 0.107 0.725 **
When compared to plans for a four-year degree or
higher: (Ed_4yr)
Undecided/other/no response (Ed_undec) -0.06 0.138 0.942
Aspire to a high school diploma or less (Ed_HS ) -0.016 0.2 0.984
Aspire to less than a two-year degree (Ed_lt2yr) -0.156 0.205 0.856
Aspire to a two-year degree (Ed_2yr) 0.183 0.154 1.201
49
When compared to students with a B cumulative
college GPA: (CGPA_B)
A (CGPA_A) 0.032 0.169 1.032
C (CGPA_C) -0.403 0.118 0.668 **
Below C (CGPA_CLO) -1.591 0.104 0.204 ****
Missing (CGPA_MIS) -1.557 0.233 0.211 ****
Students who lived on campus compared to
students who lived off campus, with parents or
elsewhere (zhouse2) 0.203 0.094 1.225 **
When compared to students who enrolled in a
public research university (En_res )
Other four-year college (En_4yr ) -0.355 0.137 0.701 **
Two-year college (En_2yr) -1.394 0.152 0.248 ****
When compared to students who received financial
aid but were not Scholars: (Aid_othe)
Aid recipients who also got Twenty-first Century
scholarships (Aid_21) 0.015 0.131 1.015
****p<0.001, ***p<0.01, **p<0.05, *p<0.10
N of cases= 2,931?
-2 Log likelihood= 3,372.811
Chi-square= 687.338**
Nagelkerke= 0.279
?236 cases were excluded from the regression due to missing
values.
50
Table 3.3 Descriptive statistics for variables in the Multinomial Regression for
Enrollment Outcomes for Low-income, Financial Aid Recipients:
Indiana's Public Colleges 1999-2003
Variable Values
Number of
cases Percentage
Male and unknown 1,338 42.2% Gender
Female 1,829 57.8%
Missing or prefer not to answer 535 16.9%
Other minority 201 6.3%
African American 435 13.7%
Ethnicity
White 1,996 63.0%
Parent(s) went to college 2,461 77.7% Parental education level
No college for parent(s) or unknown
parent education
706 22.3%
Live with one parent 988 31.2%
Do not live with parent 152 4.8%
Who the student lived with
in ninth grade
Live with two parents or unknown
living situation
2,027 64.0%
Spanish or other language 26 0.8% Main language spoken in
the home English or unknown 3,141 99.2%
Missing/did not answer 206 6.5%
Mostly As 263 8.3%
Mixed Bs and Cs and Mostly Cs 1,081 34.1%
Mixed Cs and Ds or lower 244 7.7%
Grade point average in
ninth grade
Mixed As and Bs and Mostly Bs 1,373 43.4%
Low percentage of honors graduates 2,370 79.6% Percentage of graduates
in the student's high
school who earned an
honors diploma High percentage of honors graduates 609 20.4%
Urban 650 20.5%
Rural 617 19.5%
Locale of high school
attended
Urban fringe, town and unknown 1,900 60.0%
Undecided/other/no response 540 17.1%
Aspire to a high school diploma or less 177 5.6%
Aspire to less than a two-year
degree
148 4.7%
Aspire to a two-year degree 271 8.6%
Student's aspirations in
ninth grade for education
completion
Aspire to a four-year degree or
higher
2,031 64.1%
Relatively low minority concentration 2,115 70.7% Percentage of students in
the student's high school
who are minorities
Relatively High minority
concentration
878 29.3%
Low percentage of students 2,015 66.5% Percentage of students in
the student's school who
receive free or reducedprice
lunches High percentage of students 1,015 33.5%
51
Self-supporting 355 11.2% Dependency status
Dependent on parents or indeterminate 2,812 88.8%
A 225 7.1%
B 1,113 35.1%
C 557 17.6%
Below C 1,111 35.1%
College cumulative grade
point average
Missing 161 5.1%
On campus 1,212 38.3% Living situation while in
college Live off campus, with parents or elsewhere 1,955 61.7%
Other four-year college 1,631 51.5%
Two-year college 1,114 35.2%
Institutional type
Research university 422 13.3%
21st Century Scholar with aid 391 12.3% Aid package
Other aid recipient 2,776 87.7%
Nonpersister, no degree 1,458 46.0%
Persister, no degree 1,148 36.2%
Two-year degree 257 8.1%
Enrollment outcomes
Four-year degree, including those with Both
Two- and Four-Year Degrees
304 9.6%
52
Table 3.4, Part A. Multinomial Regression Analysis of Variables Associated With
Completion of Four-year Degree, Including Those with Both Two- and
Four-Year Degrees for Low-income, Financial Aid Recipients:
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio Sig.
Female compared to male and unknown (zsex3) -0.175 0.165 0.839
When compared to Whites: (zxsethn2)
Missing or prefer not to answer 0.582 0.245 1.789 **
Other minorities -0.686 0.414 0.504 *
African American 0.33 0.327 1.392
Parents went to college compared to no or
unknown parent higher education (zxfirst2)
-0.005 0.185 0.995
When compared to living with two parents or
unknown living situation (zxslive2)
Live with one parent 0.022 0.181 1.022
Do not live with parent -0.147 0.404 0.863
Spanish or other language spoken at home
compared to English or unknown language
spoken at home (zxlangu2)
-0.535 1.112 0.586
When compared to Mixed As and Bs and Mostly
Bs GPA in the 9th Grade (zxgpa2)
Missing/did not answer -0.126 0.433 0.881
Mostly As 0.546 0.236 1.726 **
Mixed Bs and Cs and Mostly Cs 0.02 0.206 1.02
Mixed Cs and Ds or lower 0.068 0.408 1.07
Self-supporting students when compared to
students dependent on parents or indeterminate
dependency status (depend2)
-0.163 0.303 0.85
When compared to suburban/town/unknown:
(zlocale2)
Urban school locale 0.42 0.238 1.522 *
Rural school locale -0.247 0.204 0.781
Low percentage of students in the student's
school who receive free or reduced-price
lunches (when compared to high percentage)
(zlunch)
0.219 0.223 1.245
Low percentage minority students in high
school (compared to high percentage) (zminor)
0.54 0.273 1.716 **
Low percentage Honors diploma grads in high
school (when compared to high percentage)
(zhonor)
0.202 0.19 1.224
When compared to plans for a four-year degree
or higher: (zedexp2)
Undecided/other/no response -0.25 0.28 0.779
Aspire to a high school diploma or less -0.074 0.409 0.929
Aspire to less than a two-year degree -0.571 0.486 0.565
Aspire to a two-year degree -0.211 0.395 0.81
53
When compared to students with a B cumulative
college GPA: (zcumgpa)
A 0.661 0.229 1.937 **
C -0.829 0.206 0.437 ****
Below C -3.955 0.465 0.019 ****
Missing -21.866 0 3E-
10
Students who lived on campus compared to
students who lived off campus, with parents or
elsewhere (zhousing)
0.743 0.172 2.102 ****
When compared to students who enrolled in a
public research university (zenroll)
Other four-year college -0.678 0.192 0.508 ****
Two-year college -4.226 0.538 0.015 ****
When compared to students who received
financial aid but were not Scholars: (zaid4)
Aid recipients who also got Twenty-first Century
scholarship
-0.012 0.238 0.988
Dependent variable: enrollment outcome by 2003
****p<0.001, ***p<0.01, **p<0.05, *p<0.10
N of cases= 2,931?
-2 Log likelihood=4,996.503
Chi-square=273.823
McFadden=0.181
Nagelkerke=0.360
?236 cases were excluded from the regression due to missing
values.
54
Table 3.4, Part B. Multinomial Regression Analysis of Variables Associated With
Completion of Two-Year Degree for Low-income, Financial Aid Recipients:
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio Sig.
Female compared to male and unknown (zsex3) -0.24 0.217 0.787
When compared to Whites: (zxsethn2)
Missing or prefer not to answer 0.524 0.317 1.689 *
Other minorities 0.55 0.349 1.732
African American 0.045 0.425 1.046
Parents went to college compared to no or
unknown parent higher education (zxfirst2)
-0.56 0.306 0.571 *
When compared to living with two parents or
unknown living situation (zxslive2)
Live with one parent 0.523 0.221 1.688 **
Do not live with parent -0.128 0.552 0.88
Spanish or other language spoken at home
compared to English or unknown language
spoken at home (zxlangu2)
-0.045 1.086 0.956
When compared to Mixed As and Bs and Mostly
Bs GPA in the 9th Grade (zxgpa2)
Missing/did not answer -0.751 0.524 0.472
Mostly As -0.156 0.465 0.855
Mixed Bs and Cs and Mostly Cs -0.31 0.243 0.733
Mixed Cs and Ds or lower -0.649 0.459 0.522
Self-supporting students when compared to
students dependent on parents or indeterminate
dependency status (depend2)
0.178 0.302 1.195
When compared to suburban/town/unknown:
(zlocale2)
Urban school locale 0.111 0.317 1.117
Rural school locale -0.297 0.268 0.743
Low percentage of students in the student's
school who receive free or reduced-price
lunches (when compared to high percentage)
(zlunch)
0.315 0.295 1.37
Low percentage minority students in high
school (compared to high percentage) (zminor)
-0.123 0.347 0.884
Low percentage Honors diploma grads in high
school (when compared to high percentage)
(zhonor)
0.507 0.285 1.661 *
When compared to plans for a four-year degree
or higher: (zedexp2)
Undecided/other/no response 0.312 0.311 1.366
Aspire to a high school diploma or less 0.262 0.455 1.299
Aspire to less than a two-year degree -0.268 0.504 0.765
Aspire to a two-year degree 0.743 0.315 2.103 **
55
When compared to students with a B cumulative
college GPA: (zcumgpa)
A 0.22 0.326 1.247
C -0.908 0.3 0.403 **
Below C -2.069 0.284 0.126 ****
Missing -2.635 0.739 0.072 ****
Students who lived on campus compared to
students who lived off campus, with parents or
elsewhere (zhousing)
-0.252 0.247 0.777
When compared to students who enrolled in a
public research university (zenroll)
Other four-year college 0.486 0.503 1.626
Two-year college 0.949 0.507 2.582 *
When compared to students who received
financial aid but were not Scholars: (zaid4)
Aid recipients who also got Twenty-first Century
scholarship
0.706 0.28 2.026 **
Dependent variable: enrollment outcome by 2003
****p<0.001, ***p<0.01, **p<0.05, *p<0.10
N of cases= 2,931?
-2 Log likelihood=4,996.503
Chi-square=273.823
McFadden=0.181
Nagelkerke=0.360
?236 cases were excluded from the regression due to missing
values.
56
Table 3.4, Part C. Multinomial Regression Analysis of Variables Associated with
Persisting with No Degree for Low-income, Financial Aid Recipients:
Indiana's Public Colleges 1999-2003
Variables Coefficient
Std.
Err.
Odds
ratio
Sig.
Female compared to male and unknown (zsex3) -0.013 0.092 0.987
When compared to Whites: (zxsethn2)
Missing or prefer not to answer 0.517 0.142 1.676 ****
Other minorities -0.025 0.186 0.976
African American 0.266 0.165 1.305
Parents went to college compared to no or
unknown parent higher education (zxfirst2)
0.178 0.109 1.195
When compared to living with two parents or
unknown living situation (zxslive2)
Live with one parent 0.219 0.1 1.245 **
Do not live with parent -0.249 0.225 0.779
Spanish or other language spoken at home
compared to English or unknown language
spoken at home (zxlangu2)
0.08 0.49 1.083
When compared to Mixed As and Bs and Mostly
Bs GPA in the 9th Grade (zxgpa2)
Missing/did not answer -0.641 0.244 0.527 **
Mostly As 0.155 0.176 1.168
Mixed Bs and Cs and Mostly Cs -0.161 0.107 0.852
Mixed Cs and Ds or lower -0.352 0.197 0.703 *
Self-supporting students when compared to
students dependent on parents or indeterminate
dependency status (depend2)
-0.025 0.147 0.975
When compared to suburban/town/unknown:
(zlocale2)
Urban school locale 0.502 0.13 1.653 ****
Rural school locale -0.037 0.118 0.964
Low percentage of students in the student's
school who receive free or reduced-price
lunches (when compared to high percentage)
(zlunch)
0.283 0.127 1.327 **
Low percentage minority students in high
school (compared to high percentage) (zminor)
-0.055 0.146 0.947
Low percentage honors diploma grads in high
school (when compared to high percentage)
(zhonor)
0.33 0.115 1.391 **
When compared to plans for a four-year degree
or higher: (zedexp2)
Undecided/other/no response -0.086 0.146 0.918
Aspire to a high school diploma or less -0.039 0.212 0.962
Aspire to less than a two-year degree -0.069 0.214 0.933
Aspire to a two-year degree 0.138 0.164 1.148
57
When compared to students with a B cumulative
college GPA: (zcumgpa)
A -0.3 0.194 0.741
C -0.247 0.125 0.781 **
Below C -1.319 0.109 0.267 ****
Missing -1.27 0.241 0.281 ****
Students who lived on campus compared to
students who lived off campus, with parents or
elsewhere (zhousing)
0.156 0.099 1.168
When compared to students who enrolled in a
public research university (zenroll)
Other four-year college -0.265 0.145 0.767 *
Two-year college -1.344 0.162 0.261 ****
When compared to students who received
financial aid but were not Scholars: (zaid4)
Aid recipients who also got Twenty-first Century
scholarship
-0.055 0.139 0.946
Dependent variable: enrollment outcome by 2003
****p<0.001, ***p<0.01, **p<0.05, *p<0.10
N of cases= 2,931?
-2 Log likelihood=4,996.503
Chi-square=273.823
McFadden=0.181
Nagelkerke=0.360
?236 cases were excluded from the regression due to missing
values.
58
References
Advisory Committee on Student Financial Assistance (ACSFA). (2003). Review of NCES
research on financial aid and college participation and omitted variables and
Sample selection issues in the NCES research on financial aid and college
participation. Reports prepared for ACSFA by D. Heller & W. E. Becker.
Washington, DC: Author.
Hossler, D., J. Schmit, & Vesper, N. (1999). Going to college: How social, economic,
and educational factors influence the decisions students make. Baltimore: Johns
Hopkins University Press.
Musoba, G. D. (2004). Postsecondary encouragement for diverse students: A
reexamination of the Twenty-first Century Scholars program. In E. P. St. John
(Ed.), Readings on equal education: Vol. 19. Public policy and college access:
Investigating the federal and state roles in equalizing postsecondary opportunity
(pp. 153-180). New York: AMS Press.
St. John, E. P. (2003). Refinancing the college dream: Access, equal opportunity, and
justice for taxpayers. Baltimore, MD: Johns Hopkins University Press.
St. John, E. P., Hu, S., & Weber, J. (2000). Keeping public colleges affordable: A study
of persistence in Indiana?s public colleges and universities. Journal of Student
Financial Aid, 30(1), 21-32.
St. John, E. P., Hu, S., & Weber, J. (2001). State policy and the affordability of public
higher education: The influence of state grants on persistence in Indiana.
Research in Higher Education, 42, 401-428.
St. John, E. P., Musoba, G. D., & Chung, C. G. (2004). Academic access: The impact of
state education policies. In E. P. St. John (Ed.), Readings on equal education: Vol.
19. Public policy and college access: Investigating the federal and state roles in
equalizing postsecondary opportunity (pp. 131-151). New York: AMS Press.
St. John, E. P., Musoba, G. D., & Chung, C. G. (2004). Academic preparation and
college success: Analyses of Indiana?s 2000 high school class. IPAS Research
Report #04-03. Available online at http://www.indiana.edu/~ipas1/
St. John, E. P., Musoba, G. D., & Chung, C. G. (2004, July 21). Academic preparation
and college success: Analyses of Indiana?s 2000 high school class. [Report
prepared for the Indiana Commission on Higher Education and the Lumina
Foundation for Education.]. Bloomington, IN: Indiana Project on Academic
Success.
St. John, E. P., Musoba, G. D., Simmons, A. B., & Chung, C. G. (2002). Meeting the
access challenge: Indiana?s Twenty-first Century Scholars program. [New
Agenda Series, Vol. 4, No. 4]. Indianapolis: Lumina Foundation for Education.
St. John, E. P., Musoba, G. D., & Simmons, A. B., Chung, C. G., Schmit, J., & Peng, C.
J. (2004). Meeting the access challenge: An examination of Indiana?s Twenty-first
Century Scholars program. Research in Higher Education, 45(8).
59
Endnotes
1. While six-year degree attainment is becoming the minimum standard, four-year
degree attainment should still be the goal.
2. The development of longitudinal databases that track students from high school
through college requires coordination across data systems. The database
developed for the study of the 1999 cohort used information collected from a
survey on ninth-graders, along with college records and student aid applications
(see Appendix A). We did not have data on high school courses, and we used high
school information on curriculum as a statistical control for the high school level.
3. One of the major challenges facing Indiana and most states has been to place
college-preparatory curricula in all high schools. Urban and rural schools were
slower to fully implement these curriculum options than were other schools, and
course selection is still limited in these schools.
4. As of October 2004, the Indiana Career and Postsecondary Advancement Center
(ICPAC) is now officially known as the Learn More Resource Center, Indiana?s
Pre?K to College Connection.
5. With the recent decisions to make Core 40 the default diploma in Indiana, a
substantially larger percentage of low-income students should have the benefits of
this curriculum. There is a risk that raising graduation requirements in areas such
as math and language, an artifact of requiring a Core 40 for the standard diploma,
could reduce high school graduation rates (St. John, Musoba, Chung, & Simmons,
2004). However, the preparation benefits of this reform may outweigh these
potential risks. This is an issue that merits further study and debate.
6. The response rate to this survey was 80 percent. Further details on data sources
can be found in Appendix 1.
7. These studies had information on whether students applied for aid, whether they
requested to have aid information sent to an out-of-state institution, whether they
enrolled in a public college in state, and whether they applied for aid and received
aid for a private in-state institution (St. John, Musoba, Simmons, & Chung, 2002).
Students who applied for aid out of state and did not enroll in an in-state college
were probably out-of-state students.
8. These approximates are from Meeting the Access Challenge: Indiana?s Twentyfirst
Century Scholars Program (2002). If a student applied for or received state
aid but was not enrolled in a public institution, (s)he was assumed to be enrolled
in a private in-state institution. If a student applied for aid with the intent to attend
an out-of-state institution, (s)he was assumed to enroll in an out-of-state
institution.
9. The standard error for the variable ?missing grades? was substantial (see Table
2.4, Parts A & B). This is largely an artifact of the practice of not reporting grades
for students who dropped before the term was ended and grades were reported.
Our aim in these analyses was to retain all cases and note all missing cases. Given
this goal, this standard error term could not be avoided.
10. This model specification resulted in a better fit, compared with the first
multinomial logistic regression performed on all students: McFadden R2 has
60
improved from 0.16 to 0.18. This improved fit gives more credence to our
findings of significance for the coefficient on Scholars.
61
About the authors
Edward P. St. John is the Aglo D. Henderson Collegiate Professor of Education at the
Center for the Study of Higher and Postsecondary Education, University of Michigan.
His research focuses on the impact of education and finance policies in K-12 and higher
education. He is author of Refinancing the College Dream: Access, Equal Opportunity,
and Justice for Taxpayers and is series editor of Readings on Equal Education.
Jacob P. K. Gross is a doctoral student in educational leadership and policy studies at
Indiana University. His research interests include linkages between access policies and
economic globalization, critical whiteness, multicultural education, the scholarship of
teaching and learning, and persistence of low-income and historically underrepresented
students in postsecondary education.
Glenda D. Musoba holds a doctorate in Higher Education and Student Affairs from
Indiana University. She is the Associate Director/Policy Analyst for the Indiana Project
on Academic Success. Her research interests include state and federal education policy
and finance, access, equity and student persistence in higher education, and K-12 school
reform.
Anna S. Chung is a doctoral candidate in economics at Indiana University and is a
research associate for the Indiana Project on Academic Success. Her dissertation research
on students in for-profit postsecondary institutions has been supported by grants from the
American Educational Research Association and the National Association of Student
Financial Aid Administrators. She was also awarded the Cameron Fincher Fellowship by
the Association for Institutional Research for the best dissertation proposal of the year.
62
Publication information
Lumina Foundation for Education, a private, independent foundation, strives to
help people achieve their potential by expanding access and success in education beyond
high school. Through research, grants for innovative programs and communication
initiatives, Lumina Foundation addresses issues surrounding financial access and
educational retention and degree or certificate attainment ? particularly among
underserved student groups, including adult learners.
Focusing on these areas, the Foundation frames issues and explores new solutions
through fact-based research. Because we strive to be a credible and objective source of
information on issues affecting higher education, Lumina Foundation encourages original
sponsored research. The results of that research, and therefore the content of this and
other Lumina Foundation publications, do not necessarily represent the views of the
Foundation or its employees.
Believing that published research may have the longest-term impact on higher
education, the Foundation publishes and disseminates articles, research reports and
books. We prefer topics and approaches that are more practical than theoretical, and
which emphasize pragmatic tools that will assist institutions and public policy-makers.
Additional information about the Foundation?s research and grant-making programs may
be obtained from:
Leah Meyer Austin
Senior Vice President for Research and Programs
laustin@luminafoundation.org
Additional information about Lumina Foundation programs and communications may be
obtained from:
Susan O. Conner
Executive Vice President for Impact Strategy
sconner@luminafoundation.org
Lumina Foundation for Education
P. O. Box 1806
Indianapolis, IN 46206-1806
800-834-5756
www.luminafoundation.org
? 2005 Lumina Foundation for Education, Inc.