Student Persistence: Effects of Need-based Financial Aid and Merit-Based Financial Aid
Sunny Li Kwadwo Owusu-Aduemiri Florida A&M University
AIR Annual Conference June 2, 2009
Overview of Presentation • • • •
Introduction Background Purpose of the study Methodology • Data Source • Variables • Analysis
• Results • Findings and conclusions
Introduction What Matters in Persistence ? Today, most postsecondary institutions are under heavy pressure from federal and local governments to increase their graduation and retention rates in a way of fulfilling accountability to the public. An yet, most public institutions are experiencing high drop-out rates, leading to less persistence, What are the possible factors related to persistence ? Does Financial Aid affect student persistence ?
Introduction • Since The Higher Education Act of 1965, most financial aid programs are need-based, which is designed for students with low family income to reduce the financial burden associated with college education. Many programs have been established to provide need-based funding for students from low socioeconomic backgrounds, and these initiatives have assisted in increasing access to higher education for needy students. • The effects of need-based aid on college persistence have been studied extensively, in part because need-based aid has been the dominant form of financial assistance in the recent past.
Introduction cont’d • As tuition has increased over the years, states have shifted financial aid away from need-based financial aid in responding to political pressure from the middle class (Ehrenberg, 2005). At the same time college compete for the best students they can get, merit-based aid has stand out as an effective way to recruit students with strong academic performance (Duffy & Goldberg, 1998; McPherson & Shapiro, 1998).
Introduction Cont’d • While some policy analysts remain skeptical about the potential benefits of merit-based aid, others view it as part of an array of financial aid policies aimed at achieving the broad goals and objectives of higher education. • The inception of Georgia’s HOPE Scholarship program in 1993 started the trend towards meritbased aid. The rational behind most state meritbased programs is to encourage and reward academic work and promote college access within a state.
Introduction cont’d • There is an increasing concern among researchers and practitioners that resources historically assisting students from low socioeconomic backgrounds are now being used to bid for academically and financially able students (McPherson and Schapiro, 1991, Heller and Marin, 2002). Evidence of a greater reliance on merit-based versus need-based aid at fouryear public and private universities was also provided in research (e.g., McPherson and Schapiro, 1994).
Literature • Several recent studies have added empirical evidence to the merit versus need-based debate in financial aid. • Henry & Rubenste’s (2002) analysis on HOPE Scholarship Program suggested that the merit-based financial aid provided by the HOPE Scholarship Program has improved both the quality of education for collegebound students in Georgia and the equity of educational outcomes across groups. • Dynarski (2000) and Cornwell, Mustard, & Sridhar (2001) have found that Georgia’s HOPE Scholarship has increased student access to college in terms of collegegoing rates or college enrollment rates.
Literature Cont’d • The Lumina Foundation for Education's 2004 study found that need-based aid plays a bigger role in influencing high school graduates to go to college than a number of other factors, including tuition. • St. John & Hu (2004) reported that the Washington State Achievers Program, needbased, strongly improves participating students' odds of attending college. The authors' suggest that need-based programs are more effective at raising college-going rates than merit-based grants.
Literature Cont’d • Several other research has been conducted on the effects of the these two types of financial aid on college choice, enrollment, and student success (e.g. Dynarski, 2002; Turner & Bound, 2002; Alon, 2005; Bettinger, 2004; Dynarski, 2003; Heller, 1997; Leslie & Brinkman, 1987; Heller & Marin, 2002; Singell,2004; Binder & Ganderton, 2002; Cornwell, 2005; Cornwell & Mustard, 2002). • The results have been somewhat mixed and inconclusive, but the general trend seems to be that need-based financial promotes access and equity while merit-based financial aid promotes education quality.
Purpose of the Study To identify what factors play important roles in student persistence in a State University System. Examine the impact of the two different forms of financial aid (need-based and merit-based) on persistence
Methodology: Data Source This study uses 11 public universities in the State University System of Florida. The data include: Ö Ö
New FTIC students enrolled in Fall 2006, and Who applied for financial aid.
The data were extracted from the following sources: Ö Ö Ö Ö
Student Financial Aid File, Student Data Course File and Admissions File Each university submits thes files to the Florida Board of Governors on an annual or a term basis.
Metholdogy: Variables Independent variables
Gender Race Age Type of institution Family income High school GPA SAT score College GPA Need-based amount Merit-based amount
Dedependent variables Persistence Persisted: Returned or graduated Not persisted: Not returned in and not graduated by the next year
Methodology: Variables
Gender
Race
Male: Uncoded
Family Income
per $1000
Female: 1
High School GPA
GPA
White: Uncoded
SAT
SAT Score
Asian: 1, 0
College GPA
GPA
Black:1,0
Type of Institution Research University: Uncoded
Hispanic: 1, 0 American Indian: 1, 0
Non Research University:1, 0
Non-Resident Alien: 1, 0
Age
Not Reported:1, 0
Need-Based Amount
per $1000
Years Old
Merit-Based Amount
per $1000
Methodology: Analysis ÖLogistic regression is used to examine the relationship between financial aid and student persistence.
Results: Descriptive Descriptive Analysis Not Persisted Variable
Race
Asian
3.14%
6.50%
5.91%
Black
20.86%
20.21%
20.32%
Hispanic
18.28%
20.64%
20.23%
American Indian
0.61%
0.41%
0.44%
Non-Resident Alien
0.15%
0.24%
0.22%
55.34%
50.10%
51.02%
1.62%
1.91%
1.86%
Female
55.59%
58.73%
58.18%
Male
44.41%
41.25%
41.80%
Non Research University
16.20%
12.41%
13.07%
Research University
83.80%
87.59%
86.93%
Not Reported
Type of Institution
All
Percent/Mean
White
Gender
Persisted
Results: Descriptive Cont’d Descriptive Analysis, Cont’d Not Persisted Variable Age Family Income High School GPA SAT College GPA Need-Based Amount Merit-Based Amount N
Persisted
All
Percent/Mean 20.76 42,720.14 3.38 1,091.91 2.12
20.71
20.72
42,698.34 42,702. 17 3.64
3.59
1,126.09 1,120.0 9 2.98
2.83
2,833.66
3,346.40 3,256.3 3
3,918.76
5,010.90 4,819.0 5
1,975
9,268
11,243
Results: Regression Variable
B
Sig. 0.179***
0.002
Asian
-0.786***
0.000
Black
-0.451***
0.000
Hispanic
-0.274***
0.000
0.063
0.861
Non-Resident Alien
-0.681
0.314
Not Reported
-0.139
0.513
Gender
Female
Race
American Indian
Age
Years Old
-0.092*
0.018
Family Income
per $1000
0.001
0.389
0.003***
0.000
0.000
0.682
College GPA
0.929***
0.000
University Type
0.231***
0.004
High School GPA SAT
Need-Based Amount
Per $1000
0.066***
0.000
Merit-Based Amount
Per $1000
0.039***
0.000
1.305
0.271
Constant N *p