Photo: AP Photo/Philadelphia Inquirer, Ed Hille
With the number of well-paying jobs open to those without college degrees becoming scarcer by the day, policymakers have adopted an ambitious goal to increase the number of Americans with college credentials to 60 percent by 2025. As of 2016, that rate stood at just 45 percent. To increase the number of Americans with degrees by 33 percent in less than a decade, colleges will have to get much better at serving the kinds of students who have historically fallen through the cracks, and they will have to do so fast to meet that target.
Graduation rates for poor students haven’t budged much in the past four decades. Students coming from the bottom half of the family-income scale, that is hailing from families with incomes below about $65,000 per year, make up just 23 percent of Americans who receive Bachelor’s degrees by their 25th birthday, according to analysis done by the Pell Institute and the University of Pennsylvania Alliance for Higher Education and Democracy. That figure is down from 28 percent in 1970, meaning that much of the gains in college completion rates in recent decades have come from increases in the numbers of middle- and upper-class students graduating. Peter R. Jones, senior vice provost for undergraduate studies at Temple University, and Paul Tough, author of the upcoming book, “Helping Children Succeed: What Works and Why,” presented some promising solutions this month during a session at the Education Writers Association’s National Seminar at Boston University.
Jones discussed programs at Temple that provide financial resources to students the university thinks are at risk of dropping out, while Tough discussed the power of just telling students they belong in college—a potentially far cheaper solution to the college dropout crisis.
Temple University, a public research university in Philadelphia, has bucked the national trends. Despite declining funding and an increase in the number of students receiving Pell grants (a federal program that assists undergraduates from low-income backgrounds) the school’s graduation rate is up. Jones attributes that success to a philosophical change among the university’s administrators.
“We had all of these high-impact programs, but there were no students in the library, or in the writing center, or in the career center,” said Jones. “We put a lot of effort into providing resources, and as a result we have everything every other institution has. But with the type of students we have, they don’t realize they are high-needs, high-risk. We used to sit and wait. That’s changed over the last decade; we now do intrusive, or even aggressive, advising.”
It’s a lot easier to provide that kind of advising at a small liberal-arts college, but simple math means that Temple couldn’t possibly provide that kind of service to all of its 28,000 undergraduates. Temple’s solution was to turn to big data. Jones, a former criminologist, used his background tapping large datasets to create predictive models that gauged the likelihood of a person reoffending, to help Temple build a statistical program that predicts which students are at the highest risk of dropping out. The university then provides those students with increased support.
Temple’s initiative looks a lot like a highly praised program at Georgia State University, another large urban university, which is similarly crunching data to pinpoint and support students on the verge of quitting. As part of that effort, Georgia State hands out micro-grants to students to help with the kinds of minor financial impediments that have ended countless poor students’ college educations.
Through their modeling, the university has uncovered a few unexpected patterns. They learned that the students who were most likely to drop out aren’t the poorest, those who receive full Pell grants, but rather the middle class poor who only receive marginal Pell awards. They also found that in terms of predictive value, a mother’s education level matters far more than a father’s. Additionally, students with four years of foreign-language classes in high school are far less likely to drop out than their peers without that experience. Jones brought up those figures to show the power of colleges drilling down into reams of data they already possess to predict success among their students. While those trends drive Temple’s model, different factors might be at play on other campuses.
Temple has seen the proportion of its students who return for their sophomore year increase from 70 percent in 2001 to 82 percent in 2014. They’ve seen 4-year graduation rates climb from 20 percent to 44 percent and six-year rates up from 59 percent to 70 percent—that’s 11 points higher than the national average. But supporting these students doesn’t come cheap. Just one program, which offers students with the highest financial need $4,000 in exchange for promising not to work off campus for more than 15 hours a week, costs the university $8 million annually. Still, Jones said the investment is worth it because studies show students who work more than 15 hours a week are less likely to graduate.
While author Paul Tough, a two-time college dropout himself, acknowledges the role that surface-level problems like money and academic preparation play in students’ dropping out of college, he thinks that the science behind small psychological interventions – such as short surveys that have low-income, black or Latino students confront their anxieties or worries about their place on campus – is too often ignored.
“Giving college freshmen messages about belonging may not be the most intuitive answer, but it works,” said Tough. “Some worry that by suggesting this, we’re saying that the problems of low-income and first-generation students are all in their minds, and that just sitting them in front of a laptop will fix it. But that’s not what it means.”
“An earlier generation of programs and policies just said if students don’t have enough money, give them more money; if students aren’t being admitted to the right colleges, admit them,” he added. “But those programs ignored that these students are often living more complicated lives than higher-income teenagers. They are asking themselves, ‘What does this all mean? What’s my place in the world? How important is money? How do I square my place here? How does it relate to what I come from?’”
Tough said questions like these often run through the minds of students who are part of demographic groups that are statistically less likely to succeed in college, particularly on campuses where more students hail from wealthier homes.
These psychological interventions, Tough contends, have the clearest positive effect on college retention. That’s not to say supporters of this research think other interventions aren’t needed, but rather that the complex psychologies of students attending college has been overlooked, many have said. And, they add the interventions they’re exploring cost much less than other efforts to boost completion rates among underrepresented student groups.
So, one might ask, why should Temple spend millions of dollars on their programs? Jones is convinced that the initiatives are paying for themselves.
“On balance, I think we are probably making money,” said Jones. “It depends how you calculate the value of preventing a student from dropping out. But we know there’s a financial benefit. Five years ago, we just retained 4,300 students from freshman to sophomore year. This year it was 5,000, and the revenue from 700 more students is covering those costs.”
Beyond dollars and cents, Jones says that urban universities owe it to their students.
“We are enrolling students who beat the odds at Philadelphia schools, Washington schools, New York schools. We have to do whatever we can to get them through this last leg.”