News, Views and Careers for All of Higher Education
May 29
It is by now well-established that the high profile and expensive merit-based financial aid programs that numerous states have established to keep their best and brightest in college within state borders are far from the panacea their supporters envisioned. While the programs have often accomplished the goal of encouraging top-notch high school students to attend local colleges and making college more affordable for state residents, they have been criticized for disproportionately favoring higher-income students over those from low-income backgrounds and doing relatively little to encourage students who might not otherwise have gone to college to do so.
A study presented this week at the annual forum of the Association for Institutional Research suggests that, at least in one case, a state merit-based financial aid program may be working directly at odds with another priority that is near the top of concerns of most state and federal policy makers and educators: increasing the flow of Americans into scientific and technological fields.
The study, by Shouping Hu, associate professor of higher education at Florida State University, looks at his state’s “Bright Futures” program, which is one of numerous state programs designed in the image of Georgia’s Hope Scholarship Program, the first of its kind. Bright Futures, the second largest such program in the country, provides full-tuition scholarships at public colleges and comparably sized grants to private institutions to students who achieve certain minimum grade point averages in high school, and requires recipients to keep their college GPAs at certain minimum levels to sustain their awards.
Using Florida’s Education Data Warehouse, which is among the most inclusive data systems in the country for tracking the flow of students throughout a state’s educational system and into its work force, Hu examined the distribution of enrollments in various college disciplines before and after Bright Futures took effect in 1997.
What he found is that in 1995 and 1996, the two years before Bright Futures took effect, 47.5 percent of students who enrolled in degree programs at Florida’s public colleges did so in science, technology, engineering or mathematics (STEM) disciplines. In 1998 and 1999, the two years after Bright Futures took effect, 38.5 percent did, and the numbers appeared to be dropping, from 39.2 percent in 1998 to 37.7 percent in 1999.
Recipients of the Bright Futures scholarships were more likely than other students to enroll in STEM fields; in 1999, 29.3 percent of non-recipients of the merit-based scholarships enrolled in scientific and technological disciplines, compared to 34.2 and 45 percent of students who received the two types of Bright Futures grants, known as Florida Academic Scholars and Florida Medallion Scholars awards. But even the Bright Futures recipients were less likely to enter STEM fields than the average student was before the program began.
What explains the decrease in enrollment in science and math fields? “One plausible explanation,” Hu writes, is that students may have sought to “bump up” their college grades to try to qualify for, or increase the size of, their merit awards. “That is, merit-based financial aid using college GPA as a criterion for renewal could provide incentives for students not to choose degree programs in science and engineering” – which are generally seen as more difficult, Hu notes – “so that they have a better chance to qualify for the merit-based financial aid.”
Given the intense concern among state and federal policy makers about a perceived undersupply of American scientists and engineers, which has prompted significant new federal financial aid programs and many efforts at the state level, “some modifications of the current merit aid programs may be warranted,” Hu concludes.
When 84% Isn’t Good Enough
Many a public university might rest on its laurels when 84 percent of its undergraduates earn degrees within six years. But while that graduation rate puts Pennsylvania State University in the upper tier of public universities, its institutional researchers have been immersed in a multiyear project aimed at figuring out why one group of students — those from low-income backgrounds — are so much less likely to graduate. Only 20 percent of the students who come from the bottom quintile of family income and have low grade point averages in their first semester at Penn State go on to graduate, while comparably performing students from high-income backgrounds graduate at a rate of 36 percent and high-performing, high-income students graduate at a rate of 89 percent.
“Even though we have a really high graduation rate, we realize there’s a big disparity for lower-income kids,” says Michael J. Dooris, director of planning research and assessment at Penn State and a co-author of the paper, with Marianne Guidos, a quality and planning research associate. “There’s a tendency for some faculty to say, ‘Geez, how much better can [our graduation rate] be?’ But when you show them the data, they say, ‘Yeah, there’s a problem here for some of our students.’ “
To try to get at that problem, Dooris and his colleagues sought to compare the 20 percent of low-income students go on to graduate after struggling in their first semester with the strong majority who don’t, with the hope that the analysis might provide some clues for what Penn State might do to improve the odds for all of them. It’s not that Penn State is abandoning its efforts to try to increase access to college for low-income students — far from it, Dooris notes, it has been steadily increasing its financial aid budget — but as many college officials are concluding, “we have to focus on what happens to the kids who come here.”
The overall picture that emerges from the comparison of the survivors to those who don’t get through is not terrifically heartening, Dooris acknowledges. There are some bright spots: There is “no statistical evidence” that the 20 percent of low-income students who go on to graduate have stronger academic skills than those who do not, which means that “skills deficiencies can be overcome.” “The ones who graduate are probably going in and taking remedial English or math and it works for them,” Dooris says. While “some faculty believe that some of these kids are hopeless, when you look carefully at it, there’s no evidence of that.”
Among other characteristics, students with single parents were half as likely as peers with married parents to graduate, and students at Penn State’s main campus — where admissions standards are higher — were three times likelier than those at its many branch campuses to earn degrees. Family income was generally not a determining characteristic, suggesting that “while affordability is clearly an issue in general for students at this university, there is not much difference (other things being equal) in the chances of earning a degree between the relatively low-income and the lowest-income students.”
In terms of characteristics that appeared in the data to point the way to success, students who participated in work study and the passed most of their first semester course work seemed were more likely to graduate than were their peers. “These are clues, at least, that summer orientation programs, good advising, first-year seminars and similar mechanisms for students to successfully transition to college might be especially valuable for students who are most at risk,” the authors write — consistent, they note, with prevailing wisdom in scholarly research on student success.
So far, at least, says Dorris, the research, while useful, has not produced any magic bullets. “We’re not solving this problem by a longshot,” he acknowledges. “There’s a real problem, and to some extent, it’s a social and a cultural problem. But this kind of study suggests that if we keep working at it, we can make a difference, and we have no alternative but to do that.”
Want it on paper? Print this page.
Know someone who’d be interested? Forward this story.
Want to stay informed? Sign up for free daily news e-mail.
Advertisement
The Penn State study should be expanded to more than just one institution in order to eliminate the possibility of selection bias (e.g., students at Penn State might have different persistence characteristics than students at other types of institutions). For example, the result concerning family income might be specific to the success of Penn State’s financial aid programs in minimizing the impact of financial strength on persistence. Or it could be due to self-selection for the type of student who enrolls at Penn State.
The other data suggests the need for proactive intervention for at-risk populations throughout the college years. While there are intervention programs to help stop pipeline leakage during the elementary, middle and secondary school years, much of the effort during the postsecondary years is reactive and passive, not proactive. For example, 1-800-4-FED-AID requires the student and/or family to make a call. Nobody is calling them to see how they are doing, to intervene and provide help when they encounter problems, to nag the student to study more and party less, and so on.
It is my belief that nationwide we need a combination of a significant increase in need-based student aid (e.g., double the Pell Grant and allow the EFC to go negative) coupled with proactive intervention and counseling for at risk populations. This will eliminate two key stumbling blocks that prevent low income students from succeeding in college.
Mark Kantrowitz, Publisher at FinAid, at 9:25 am EDT on May 29, 2008
With such a deep data set available, why look only at enrollment? If the effect is due to concerns about college GPA, it would show up as a shift in majors after first-semester troubles in calculus or chemistry. Did the number of STEM graduates change as dramatically four (or six) years later?
If the effect shows up in enrollment numbers, could it reflect a change in the HS classes being taken, to raise the HS GPA, rather than knowledge of the challenges that lie ahead in college? Are HS grads really that aware of how much more work is required for most STEM majors? My experience says they aren’t. Of course, it could be that colleges were advising them to take a different path.
I am also troubled by subjective comments such as “... degree programs in science and engineering” – which are generally seen as more difficult, Hu notes. Again, with a data set as deep as is described in the article, it should be an easy matter to compare the grade distributions and mean course GPA for pre-STEM classes (like the one I teach) and those required to get into the majors (unidentified in this article) that saw increased enrollment while STEM enrollment decreased.
CCPhysicist, at 9:30 am EDT on May 29, 2008
Florida’s Bright Futures is a disastor, and the lack of political leadership in the state to change it is alarming. University presidents are not committed to changing it out of fear of losing money, the state legislature is not willing to change it because of middle class political suopport, and the Board of Governor’s is not inclined to lobby against it either. The program is here to stay.
Yet it is clear:
-Bright Futures keeps Florida tuition artifically low and hinders improvements in quality at Florida universities.
-Bridge Futures does not enhance acess to higher education for the poorest students who need it most.
-Bridge Futures benefits upper-middle class students more than anyone else.
-Bridge Futures has discouraged students from staying in STEM fields and thus fails to serve Florida’s economy, which needs more scientists, engineers and doctors.
I have no confidence in Florida Higher Education or in its leadership. Someone needs to dismantale this program before it turns the entire Florida system into a massive community college.
FSUfather, at 9:50 am EDT on May 29, 2008
In Penn States search for the cause of such gaps in success, I would encourage them to look at the work of Estela Bensimon on the equity scorecard (University of Southern California. Dr. Bensimon’s action research has identified a way for faculty to get a real handle on just where in the curriculum at-risk students are being lost. It’s very good work, and worth exploring.
Dave Longanecker
david Longanecker, President at WICHE, at 10:40 am EDT on May 29, 2008
We have read here on Inside Higher Ed about grade inflation. Just last week, we read about a professor who lost his position for flunking excessively. Grade inflation is particularly operative in the social sciences and humanities. Or so I read here. Way less so in the sciences. You can study a lot in chemistry, but come up with a B- ooops not so good. Whereas, in the soft areas, opinions count more than correct lab analyses, and so. My pre-med son was choring through Organic Chemistry—hoping for a B+ maybe, and settling for a B, his suitemates were waltzing through Communications in the New Millennium, and partying and getting A’s or A- maybe. College students are not stupid. If the gpa is what counts, take the cake courses, NOT math or science.
innocent bystander, multiple factors operating here, at 12:10 pm EDT on May 29, 2008
One of the key reasons lower income students graduate at lower rates is because a lower income student’s perception of a “good” salary, and a “good” job are different than those of higher income students. They often drop out because they find that the salaries that are available to them are quite adequate for their perceived needs. Peer group effects are also important for students at all income levels. Peer groups reinforce perceptions of what constitutes an adequate salary, working conditions, vocational advancement, and lifestyle.
Steve, at 12:10 pm EDT on May 29, 2008
The FEDW is indeed an amazing data source and could be goldmine for research purpose, and people know that. Perhaps because of the good fame of the FEDW, the FEDW staff members, who are wonderful professionals, have to respond to large volume of data requests. So it could take some time for the requester to finally get the data. Furthermore, researchers may not be aware of what elements are available and what the data structure is like. It might not be uncommon that the researcher, when got the data, feels that some additional variables are needed. I would have liked to have student GPAs semester by semester, in addition to many other things. The AIR presentation is a part of a more ambitious undertaking, including the plan to examine student pathways to baccalaureate degrees in STEM include a sequence of decisions pertaining to the initial choice of a science or engineering field, change of major field, and completion of a science and engineering baccalaureate degree. The plan is also to examine students of different gender, race/ethnicity, and students with different college grades. Further, it also plan to use cross-sectional analysis, longitudinal “before and after” comparison, and further an analysis on panel data. However, given the scope of the plan, the complexity of the data, and the limited resources (human, financial, and computing infrastructure) available, the task is much more challenging and the progress is much slower……The good thing is that it is doable……
Grade inflation is another interesting topic, and the genesis of this AIR project is from my previous work on that topic, which ends as a monograph titled “Beyond Grade Inflation: Grading Problems in Higher Education” (2005, Jossey-Bass). It also becomes a conceptual anchor for this “Bright Futures” project.
Thanks for the constructive comments and suggestions.
S. Hu, at 1:20 pm EDT on May 29, 2008
There are so many errors in the logic of these statistics that it’s not even worth arguing about. I hate it when studies are released that present statistics that have obvious immediate questions that are left unanswered. It’s as if they stopped the study at the point where it would make a good news article and then went on to something else.
John, Statistics 101, at 3:05 pm EDT on May 29, 2008
I am a graduate of Penn State, class of 2001. I was one of those lower income, single parent freshman who was one tiny statistical count in this survey.
A graduation rate of 84% is very hard to believe based on my experiences at Penn State. I started at a branch campus, where the majority of the “cash cattle” students start. The great majority of those who started with me had dropped out within 2 years, to the point where even the 20% graduation rate for this group cited in the article seems generous.
Like many large multi-campus schools, Penn State could not actually cope with a higher graduation rate. If all of those branch campus droppouts were to stay, they would overwhelm a main campus which has not enough housing or classroom space to accommodate them. In this sense, Penn State treats enrollment like an airline books seats, expecting and counting on a number to droppout, but still collecting the tuition fees.
Shawn, Penn State Graduate at Penn State, at 4:25 pm EDT on May 29, 2008
Perhaps the reason for failure to complete the program is as simple(?) as a lack of support system, as suggested by another poster. I’m not sure it should be considered the responsibility of the school to provide this, but it wouldn’t hurt to have resources for first generation college students. Another possibility that sort of goes along with this could be that first generation higher ed-ers view it as more of a “dream” than a “goal” or “plan” which also speaks to Steve’s post when presented with the easier alternative — the sufficiently paying job that does not require the degree. With little to no real support at home and/or community for achieving at this level, it’s more likely one will give up when the going gets tough.
NJ, at 4:25 pm EDT on May 29, 2008
“Grade inflation is particularly operative in the social sciences and humanities. Or so I read here. Way less so in the sciences. You can study a lot in chemistry, but come up with a B- ooops not so good. Whereas, in the soft areas, opinions count more than correct lab analyses, and so.”
I found social science and humanities courses to be graded much more rigorously than the sciences. The introductory social science courses at my big state school were all tested multiple choice, so there had to be right or wrong answers. As for the humanities, this month’s Atlantic has a great piece about how standards are kept up: http://www.theatlantic.com/doc/200806/college
Also, conflating STEM together is deceptive, TEM graduates still make decent salaries, but my understanding is that science salaries have been stagnant for decades. Even doctors are not doing as well as they once did.
The real money has always been made in business or in the higher echelons of law. The best undergraduate preparation for these careers is either a social science degree (economics), or any pre-law major. The results of this study are not surprising at all.
Steve, at 5:10 am EDT on May 30, 2008
Advertisement
or search for jobs directly.
The nation’s first university, Penn is a world-renowned leader in education, research, and innovation. Situated on a ... see job
Angelo State University’s work environment is often described as an employee community. This applies to the work culture ... see job
Everest College, a respected member of the Corinthian Colleges’ network of schools, is dedicated to helping students ... see job
Permanent Full-Time EPA position at Tri-County Community College Main Campus in Murphy, North Carolina. Position will design ... see job
Position Summary:
The nation’s first university, Penn is a world-renowned leader in education, research, and innovation. Situated on a ... see job
Job Description: Ithaca College invites applications for the position of Registrar. This position is ... see job
Join the Pack! A community with nearly 8,000 faculty and staff, and 30,000 students. NC State is one of the largest employers ... see job
Everest College, a respected member of the Corinthian Colleges’ network of schools, is dedicated to helping students ... see job
Salary: Commensurate with experience Date of appointment: October, 2008 or as negotiated Responsibilities: St. Cloud State ... see job
Male student have historically been well-represented in STEM disciplines. Female students get better grades in high school and college and one would expect they are overrepresented in programs such as ‘Bright Futures’. Could this be the result of ‘Bright Futures’ lowering the proportion of male students in the university, particularly sharp male students with lower GPAs than the ‘Bright Futures’ cutoff, a decline that would then be followed by an overall lower proportion of students in STEM courses? Any data on the proportion of females in ‘Bright Futures’ and the effect on male/female ratios?
Any Data?, at 8:00 am EDT on May 29, 2008