Community Colleges

Looking Beyond the Data to Help Students Succeed

April 22, 2015

If Santa Monica College had relied solely on data analytics to predict whether Jaime J. would succeed, the picture would have looked bleak. He was, after all, a financially struggling, first-generation Hispanic student who was juggling a job with classes. His math skills were shaky.

But there was more to the picture than that. Using a 30-minute online assessment that focuses on noncognitive skills, advisers at the two-year institution in Southern California learned that Jaime was also a conscientious student with good study habits who had long dreamed of becoming a computer engineer.

The college assigned him a success coach (the college’s dean of counseling and retention), who met with Jaime weekly to keep him motivated.

"He now has 89 transferable units and a 3.2 GPA," said Brenda Benson, the Santa Monica dean who stepped in as a coach to get a firsthand look at how the program was running. "There were times when he wanted to give up, but I wasn’t about to let him."

She was speaking here on Tuesday at the annual convention of the American Association of Community Colleges, where predictive analytics and retention strategies have been hot topics of conversation.

Everyone, it seems, is swimming in data about their students, but they’re not sure what to do with it and are wary of using it in ways that reinforce stereotypes about students’ abilities.

The program that Santa Monica has been trying out is SuccessNavigator, an assessment developed by the Educational Testing Service. Some 25,000 students on about 150 college campuses (half of them community colleges) have taken the personal survey since it was released, in July 2013. It was designed to help colleges get a more holistic picture of students’ chances of succeeding in college.

The assessment, which costs colleges $6 per student, provides feedback to students, their advisers, and instructors, on the specific hurdles students may face and links them with campus resources that can help.

"We’re really good at building predictive models" based on demographic, socioeconomic, and other characteristics that educators have no control over, Ross E. Markle, a senior research and assessment adviser for ETS’s higher-education division, said during Tuesday’s session. "The challenge is, What do we do once we know that a student is unlikely to succeed?"

What is it, specifically, about a first-generation student’s background that makes college so hard, and what factors can the college help such students change?

Study Skills and 'Grit'

While many students have weak academic skills because of language barriers or inferior secondary schools, a variety of noncognitive traits can also hurt or help.

Those include things like study habits, time management, self-confidence, and test-taking strategies. Another is "grit," a popular term in higher-education circles these days that is used to describe perseverance or resilience.

Santa Monica’s new slogan, posted on signs across the campus, is "Got Grit? Passion, Purpose, Perseverance."

Social connections are also important. Connecting with an encouraging mentor can raise a student’s morale; being told by a parent that college is a waste of time and money can be deflating.

Those are all factors that educators can help students change, unlike the fixed characteristics of ethnicity and income level that colleges often focus on, Mr. Markle said.

Homing in on the specific traits that can make or break a student is the idea behind SuccessNavigator, which is usually administered just before or early in the enrollment process.

The assessment is a series of questions like whether the student completes assignments on time or shows up for nearly every class.

Topics cover classroom and study behaviors, commitment to educational goals, management of academic stress, and connection to social resources.

Skeptics have questioned whether students would answer questions about their study habits honestly if they felt it might hurt their chances of getting admitted or avoiding remediation.

But Mr. Markle said that when administrators explain that the main purpose is to connect students with the supports they need, most are willing to confess if they tend to sleep through morning classes or plant themselves in the back row.

The goal, Mr. Markle said, is to "reach out to students before they realize they need it."

By the time they’re struggling, many at-risk students are reluctant to visit a tutor or counselor.

"Many have that impostor complex and figure, If I seek help, they’ll know I don’t belong," Mr. Markle said. Better-prepared students, by contrast, "grab every resource they can. That’s why tutoring centers are filled with more B-plus students trying to get an A than D-plus students trying to get a C."

The test can also help an adviser decide whether a student who scored near the cutoff on a career-readiness measure like the College Board’s Accuplacer tests might be able to skip remedial course work and go directly into a college-level class.

The Community College Research Center at Columbia University’s Teachers College argues that too many students who could succeed in credit classes are routed instead to long sequences of remedial classes on the basis of placement tests that are incomplete predictors of success.

A more-nuanced assessment, Ms. Benson, said, helps students like Jaime reach their potential.

Katherine Mangan writes about community colleges, completion efforts, and job training, as well as other topics in daily news. Follow her on Twitter @KatherineMangan, or email her at katherine.mangan@chronicle.com.