Session M1E 1-4244-0257-3/06/$20.00 © 2006 IEEE October 28 – 31, 2006, San Diego, CA 36 th ASEE/IEEE Frontiers in Education Conference M1E-20 Student Strategies for Protecting Merit-Based Scholarships: Grades, Courseload, and Major Choice Guili Zhang 1 , YoungKyoung Min 2 , Sharron A. Frillman 3 , Timothy J. Anderson 4 , and Matthew W. Ohland 5 1 Guili Zhang, Assistant Professor, East Carolina University, Curriculum and Instruction, East Fifth Street, Greenville, NC 27858-4353; zhangg@ecu.edu 2 YoungKyoung Min, Educational Statistics, University of Florida, Office of the Dean of Engineering, PO Box 116550, Gainesville, FL 32611; ykmin@ufl.edu 3 Sharron A. Frillman, Purdue University, Engineering Education, 400 Centennial Mall Dr, West Lafayette, IN 47907; sfrillma@purdue.edu 4 Timothy J. Anderson, University of Florida, Office of the Dean of Engineering, PO Box 116550, Gainesville, FL 32611; tim@ufl.edu 5 Matthew W. Ohland, Purdue University, Engineering Education, 400 Centennial Mall Dr, West Lafayette, IN 47907; ohland@purdue.edu Abstract – The effects of merit-based scholarships on engineering students are studied by relating the GPA, credit load, and retention rates of engineering student merit-based scholarship recipients and students who met the same SAT and high school GPA criteria but did not receive the scholarship because it was not offered at the time of their enrollment. Student record data for 10,167 engineering students at three institutions in the state of Florida from 1987 to 2002 are extracted from a nine- institution database with student records data from 1987 to 2002 compiled by the Southeastern University and College Coalition for Engineering Education (SUCCEED) and used to examine the differences between the two groups of students in terms of GPA, semester credit hours, and percent leaving engineering over time. The comparisons reveal evidence that merit-based scholarships led students to adopt three strategies for retaining their scholarships: scholarship students tended to have higher average GPAs, tended to take fewer credit hours, and were more likely to leave engineering than similar students matriculating prior to the implementation of the scholarship program. Index Terms – Enrollment management, financial aid, merit- based scholarships, retention. THE SHIFT OF AID FROM NEED-BASED TO MERIT-BASED The College Scholarship Service was created in 1954 to devise a standard formula for determining financial need. Recognizing the potential for inequities inherent in a strictly merit-based system, most institutions adopted the policy of awarding scholarships according to need [1]. Starting in the early 1990s, academic merit has gradually been displacing financial need as the basis for awarding grants and scholarships. Between 1989 and 1995, however, the number of need-based awards grew more rapidly than that of non-need awards at all four-year institutions, yet the average amount of non-need grants increased at a much faster rate than that of need awards. Merit-based awards had become a tool for recruiting students with high academic qualifications. Longanecker, however, found that increases in merit-based aid have not necessarily been at the expense of need-based aid [2, p. 32]. Georgia’s “Helping Outstanding Pupils Educationally” (HOPE) Scholarship was initiated in 1993 [3]. Though the HOPE Scholarships originally had both merit (GPA of 3.0) and need-based criteria (annual family income of $66,000 or less), the family income cap was eroded and by 1995 it was removed entirely [4]. Concerns have raised previously over the effects of the merit- aid approach due to a widely held belief that merit awards tend to go disproportionately to students who would have attended college with or without financial assistance [5]. Confusing the issue somewhat is evidence that merit-based criteria influence the selection of need-based scholarships. After noting a high correlation between socio-economic status (SES) and academic achievement, Heller and Rasmussen explored this interaction [6]. Their findings were consistent with those of McPherson and Schapiro [7], who found that low-income students with high SATs attending public institutions received on average much larger need-based grants ($1,255) than students with average ($904) or low ($565) SAT scores. A similar relationship was found to exist among attendees of private colleges. The authors found SAT scores to be strongly correlated with levels of need-based grants at both public and private institutions. PREVIOUS STUDIES OF MERIT-BASED AID Research on the effect of the HOPE Scholarship has focused on students’ course-taking patterns [3] and the factors affecting scholarship loss [4]. Among University of Georgia students, Cornwell and colleagues [3] observed a decline in completed credits of 0.8 quarter-hours (approximately one sixth of a typical five quarter-hour course) for Georgia residents as compared to non-residents. They also observed a shifting of course work from the academic year to the summer session. A later study by the same research team [8] showed that (a) resident first-year GPA increased by 0.13 point, although this effect weakened after the first year; (b) both residents and non-residents maintained lower GPAs in math and science courses; and (c) student major-changing patterns had been affected [8, p. 17]. While Cornwell et al. presume that “HOPE may induce a substitution away from [math and