Page 35 - summer18

[ Fourth Industrial Revolution ]

RETENTION: AN AGE-OLD ISSUE IN HIGHER ED

am’s research represents a new approach improvement over previous models, which
Rto an old problem, and she hopes it can relied largely on descriptive data from surveys ‘What all of this
eventually be used by the UA and other and incoming freshmen’s “academic index
universities to supplement the predictive scores,” predictions of their first-year college ultimately boils down
analytics work they are already doing. GPA based on their high school GPA, the
“Student retention is something that’s been difficulty of the courses they took in high
studied for the last 30 or 40 years, but we never school, and their SAT or ACT test scores. to is individual students
had the ability to track people’s behavior and The university now generates lists, based
movement and extract their social integration on those 800 data points, twice a semester and individual needs
patterns,” Ram says. “We have also made great and twice in the summer of the 20 percent of
strides in developing machine learning and students most at risk in each college. Those lists and how we translate
large-scale network analysis methods that help are shared with the colleges with the intent that
in analyzing such spatiotemporal data.” advisers will use them to reach out to students the big signals down into
Ram’s work comes at a time when universities who may need additional support or guidance,
nationwide, including the UA, are committing Baldasare says. individual outreach.’
more resources to harnessing data analytics in “As early as the first day of classes, even for
ways that support student success. freshmen, these predictive analytics are creating
“The kind of move that universities are highly accurate indicators that inform what
making toward predictive analytics and using we do to support students in our programs and
more data to understand the student experience practice,” she says. “The lists of students are
allows us to look earlier and more often at some delivered in the fourth week of classes because
of the variables that we can get our hands on and we know students may already be making up
ask different sorts of questions than we were able their minds about staying or leaving the UA.”
to ask before about the freshman experience,” The university also is preparing to launch
says Vincent J. Del Casino Jr., UA vice president an online dashboard where advisers can access
for academic initiatives and student success. key data and assess student risk in real time
The UA, which saw its freshman retention rate throughout the semester.
jump from 80.5 to 83.3 percent between 2016 and “We’ve really worked hard over last year to
2017, has for the past four years contracted with work more with advisers and get advisers better
outside vendor Civitas Learning for data analysis data so we can facilitate their informed action
related to retention and graduation rates. with students,” Baldasare says.
The university now uses some 800 data Predictive analytics merely provide a signal
points — related to everything from academic for underlying challenges students might be
performance to financial aid to students’ activity facing, Baldasare notes. It’s up to the advisers on
in the university’s D2L course management the ground to use that information to diagnose
system — to identify which students are most the problem and help students as best they can,
at risk for leaving. Those predictions, which do with the understanding that it will never be
not include data from Ram’s research, are about possible to retain everyone.
73 percent accurate from the first day of classes, “What all of this ultimately boils down to
with the rate of accuracy improving over time, is individual students and individual needs
says Angela Baldasare, UA assistant provost for and how we translate the big signals down into
institutional research. individual outreach,” Baldasare says.
The university’s current effort to predict
first-year student retention is a significant

SUMMER 2018 33