Bonner Student Progression Study
Bonner Student Progression Study Approach
The Bonner Student Progression Study (“Progression Study”) sought to leverage data that institutions collect and report (to the The Integrated Postsecondary Education Data System or IPEDS) for analysis. The study focused on understanding the gains of students involved in the Bonner Program when compared to peer students who do not participate in the program. This analysis can provide a “proof in concept” for how the cohort-based four year model acts as an intervention. The study aimed to further the Foundation’s understanding of a number of important questions posited by many over the years including our program staff, faculty, students, our presidents and other administrators, national and community-based partners, and Board of Trustees:
What are the differences between Bonners and other students (non-Bonner Program participants) for key progression outcomes such as term-based GPA, attempted and earned hours, and retention?
What differences exist within and between schools for common indicators such as time to degree, degree GPA, and graduation attainment?
What differences exist for progression and graduation outcomes by group (such as income level, key pre-college aptitude ranges, socio-demographic factors, etc.)? Do we see variances for these groups between Bonners and other students?
Seven institutions – Berry College, College of Saint Benedict, Guilford College, Stetson University, The College of New Jersey, Ursinus College, and Wofford College – provided data for the pilot. The Foundation seeks to carry out the similar analysis with other colleges and universities in the network building upon what we learn from the initial study.
Bonner Progression Study Methodology
With help from Institutional Research offices and staff, we obtained student record data that permitted an assessment of differences between Bonners and other non-participating program students for the following:
First through Fourth Term GPAs obtained by the students
First through Fourth Term Attempted and Earned Hours
1st, 2nd and 3rd Term retention and attrition outcomes
Degree attainment: Degree Attainment within 150% of time (y/n)
Degree cumulative GPA
The available data allowed the research team an opportunity to pursue two major questions:
Do Bonner Scholars and/or Bonner Leaders perform at a level congruent or above the general population of the institution (“non-Bonners”)?
Do students in the program who come from groups historically under- represented (racial/ethnic subgroups) in higher education and/or from lower socio- economic levels (e.g., Pell eligible) perform at a level congruent or above like-students in the general population or students who do not come from these groups within the program or the campus broadly?
HYPOTHESES: ACCESS vs LEADERSHIP ORIENTATION
As a program designed to increase college access, it is reasonable to suggest that student integration is a primary outcome of an intervention that provides educational support and curricular and co-curricular interventions to students from low-income backgrounds. Bonner Scholars and Leaders are part of a population of students who are, as low-income students, at-risk for not attending or completing college because of the relationship of socio-economic status and lower pre- college academic performance as measured by entrance assessments such as SAT, ACT, and high school cumulative grade point average, and rank.
A common goal for access programs is to support students to perform at levels congruent with the general population within the institutional environment. On the other hand, the Bonner Program has also functioned as a leadership focused program (especially for Bonner Leader Programs that do not have the same financial aid obligations as the Bonner Scholar Programs). The goal of leadership programs is often to support students to achieve academic and levels of status (leadership roles, meritorious recognition, etc.) beyond the general population. Although all Bonner Programs across the network seek to educate and graduate their students, they vary in the profile, focus, and attainment expectations of program participants.
With these considerations in mind, the data gathered from the institutions revealed some interesting differences between the programs at colleges and universities in this study. After assessing pre-college variables and financial aid profile of samples at each school, this taxonomy was used to context findings for the schools.
For the first-term GPA variable, Bonner Program participation overall (i.e., all schools combined) exhibited a marginal significance (p<.06) level for first term GPA mean as compared to those not participating. This finding is very important from the vantage point of understanding positive program outcomes. It suggests that participants generally performed at levels similar to non-program participants and even slightly outperformed these other students at a marginally statistically meaningful level. Integration (and more) is being accomplished by students in the first term of the experience. Additionally, 5 of the 7 Bonner programs (the “access” schools) exhibited statistically significant and much lower baseline variances from the non-Bonners. Specifically, they exhibited -
Higher financial need as measured by Adjusted Gross Income (“AGI”) and Estimated Family Contribution (“EFC”);
Higher subgroup membership (Pell, racial/ethnic minorities) associated with progression risk (5 of 7); and
Lower scores on pre-college measures (SAT, ACT, HSGPA) associated with the effective prediction of progression outcomes (4 of 7 schools).
Given these baselines for key covariates such as pre-college SAT and ACT scores, High School GPA, and financial need, the “integration” of Bonner students and the similar performance on the first term GPA is a very important finding given it is often viewed as the most important post-matriculation indicator of retention and graduation along with first term earned hours.
This extraordinary finding for the many Bonner students coming to college with lower baseline scores and higher financial need likely speaks to the extra support, advising, and peer influence during the first term for Bonner program participants. For the two programs classified as leadership focused (Ursinus College and the College of Saint Benedict), the Bonners in these programs exceeded the first term GPA performance of those not participating in the program at a threshold that was statistically significant. Additionally, findings for the subsequent term GPAs (2nd through 4th term) were similar for Bonners. Again, this suggests that the students performed within the same levels as their counterparts at those schools with an access-oriented program mission and Bonner performance extended beyond their counterparts for the term-based GPA at those schools whose programs were primarily leadership-oriented.
Another impressive finding of the study was for third-term (1st year) retention. The Bonner Program participants exhibited mean overall differences for the variable (i.e., higher overall retention rates) at a high level of significance (p<.001) with and without controlling for key covariates (e.g., SAT Composite, ACT). Although program participants generally perform at levels congruent with non-program participants for term-based GPAs as noted above which is especially a notable finding for those programs that are access-oriented. It is also true that these students receive an inordinate amount of support (financial, peer, faculty, advising) and this academic and social integration likely strengthens their ability to persist above classmates. We find the mean difference findings are further reinforced by looking at bivariate correlations where the relationships are very significant (p<.001) for the outcomes retention (1st thru 4th terms) for program participation vs. non-participation.
For key subgroups (Pell, minority-all, Hispanic, African American), we did not find statistically significant relationships for students in the program in these groups vs those assigned to these groups that were not in the program. However, Bonner Program participants generally exhibited descriptively higher rates for 3rd term retention for these subgroups (e.g., African American, Minority, Pell) as compared to non-Bonner like-subgroups across programs. Thus, the program works effectively for students of different backgrounds. In examining graduation, Bonner Program participation exhibited positive and statistically significant correlations and mean differences (p<.001) for degree attainment (1=Attained, 0=Not Attained) for all schools when the sample was combined. This means that program participation (sample combined) has a statistically meaningful relationship at a magnitude that is stronger and related with degree attainment. In other words, Bonner students have a higher association with obtaining a degree than non-Bonners (from similar backgrounds).
The Bonner Scholar Program participation within this sample of schools appears to promote integration with the campus in terms of its progression outcomes related to both term-based and degree GPA, retention, and degree attainment.
Specifically, Bonners perform at levels similar to the general population or some cases (leadership-oriented schools) exceed these levels with one campus exception.
Key subgroups (Minority, Pell, etc.) also perform at levels that exceed like-groups in non- Bonner populations (descriptively).
As noted, the majority of programs reflect distributions and means similar to the general population (i.e., non-Bonners) and this finding is particularly significant given 5 of the 7 Bonner programs exhibited statistically significant baseline variances from the general population.
Additionally, 4 of the 7 Bonner Program participants exhibited lower scores on pre-college measures (SAT, ACT, HSGPA) associated with the effective prediction of progression outcomes.
The Bonner Programs that exhibit largest difference from non-Bonners appear to be focused on selection criteria related to leadership and academic aptitude and less on access.
These positive findings suggest that additional analyses can be pursued to examine a broader array of institutions (especially access-focused Bonner Scholar Programs).