"Summer melt" -- referring to those students who commit to enroll at a given college but then don't do so -- has been the subject of much discussion in recent years. Articles have focused on the challenges that colleges face when students put down multiple deposits (which they aren't supposed to do), throwing off projections of "yield," the percentage of accepted applicants who actually enroll. Students and their families have been encouraged to play off of colleges' fears of too much summer melt to ask for more financial aid (frequently of the sort that is not based on financial need).
But there is another kind of summer melt -- one that may have much more of a societal impact than whether a middle class student ends up at College A or College B and one that affects students who wouldn't have the money to put down multiple deposits. This kind of summer melt involves low-income students who have almost made it into college, but may not make it there -- students who applied, were admitted, filled out their financial aid forms, and qualified for substantial assistance.
A paper being presented this week by Benjamin L. Castleman and Lindsay C. Page, both doctoral students in education at Harvard University, shows significant summer melt for such students, especially those planning to enroll at community colleges. The paper -- which will be presented at the annual meeting of the American Education Research Association, in New Orleans -- suggests specific actions colleges and other institutions can take to minimize this kind of melt.
Given the importance of community colleges in starting students (especially low-income and minority students) on the path toward bachelor's degrees, the paper suggests that addressing this kind of summer melt may be key to various national goals for increasing the percentage of the public with either an associate degree or a bachelor's degree.
Castleman and Page examined the patterns of college enrollment of high school students tracked in two databases. One was the ACCESS program that works with low-income students in the Boston public schools to help them get aid for college. The other was a national database -- the Education Longitudinal Study of 2002. The former, while from one large city, focused on students who in many ways are key to increasing educational attainment -- those who are low-income but have connected to a program that will find them aid. And most of the students are so low-income that most would have no expected family contribution and so would qualify for maximum levels of federal and (in many cases) institutional aid.
Over all, Castleman and Page found that 22 percent of the ACCESS students who were "college intending" at high school graduation (and who thus seemingly could enroll) melted during the summer. The figures varied significantly based on the college to which students expected to enroll: melt was 19 percent for four-year institutions and 37 percent for community colleges. The lower students' income is, the more likely they are to melt.
In looking at characteristics of the students who enroll and those who don't, the research found strong impact of parents and peers. Students who "talked frequently with their parents about college" are 7 percentage points more likely to enroll than are those who talked only infrequently with their parents. The researchers also found "strong and positive association" between the proportion of a student's friends with college plans and college enrollment. If most of a student's friends plan to enroll, the student is 14.2 percentage points more likely to enroll.
Among the possible implications of their findings is that high school counselors need to stay in touch with low-income students over the summer, to be alert for signs of melt so that intervention is possible. (The authors note that this and other ideas that cost money may be difficult for the financially strapped schools that serve many such students.) Castleman and Page write that their research suggests that the most effective counselors in the summer between high school and college will be those who already have a relationship with the students. They also suggest the use of social networking as a way to stay in touch with these students and help them deal with last-minute obstacles to enrollment.
While a number of colleges have "bridge" programs that help disadvantaged students the summer before freshman year, the authors note that, while these programs are successful, very small percentages of low-income students actually enroll in them or have the chance to do so.
The researchers are planning a random trial in Boston this summer -- with other researchers conducting similar trials in Fort Worth, Tex., and Fulton County, Ga. -- to see if specific intervention strategies can minimize the melt.
They write that they hope to find approaches that could be carried out, given that their research demonstrates that low-income high school graduates with college plans are "particularly susceptible" to having those plans fall apart.