Colleges can now figure out which students will be successful - even before classes start
Adam Berry/Getty Images
Adam Berry/Getty Images
By the end of some students' first two weeks in a college course, an analytical model can determine with 75% accuracy rate how well they'll end up doing.But should it?Advertisement
For the 2,200 students at community colleges and historically black universities, the Open Academic Analytics Initiative in 2014 - a program developed by Marist College and business analytics firm Pentaho -tracked habits like clicking on online reading materials, whether they posted to online forums, and how long they needed to complete their homework.
The data then funelled into a system designed to predict academic success, stage interventions, and push for higher graduation rates.Other colleges have widely begun to adopt similar practices. At the University of Maryland University College, where most of the 60,000 undergraduates take courses online, administrators mine data from students' activity before courses begin - like if students look through their syllabi online - and are experimenting with automatic alerts for students falling behind in their classes.
"We know the day before the course starts which students are highly unlikely to succeed," Marie Cini, the provost at UMUC, told the Chronicle of Higher Education.These practices, however, raise ethical questions about whether collecting this kind of data on students invades their privacy."We are entering a new era of data and data responsibility," Mitchell Stevens, an associate professor in Stanford University's Graduate School of Education, told the Chronicle. "Are we acting responsibly as educators? What values are we trying to pursue and preserve?" Advertisement
Stevens and a group of experts met to discuss the ethics of student data collection at a private conference this month, according to the Chronicle. The conference intended to "consider how data describing adult students might be managed in ways that enable the improvement of educational experiences, the progress of science, and the integrity of information describing human beings," according to its website.
The last time the group met in 2014, it discussed how to handle the massive amount of information generated by students taking online courses, many of whom aren't officially students at the college offering the course. This year's meeting focused on how institutions should treat their own students when it comes to data collection. Advertisement
Official findings won't be released until August, but the group provided the Chronicle with some rough conclusions. Among them:
- Data collection on students should be considered a joint venture, with all parties - students, parents, instructors, administrators - on the same page about how the information is being used.
- Data-analytics programs need to be transparent, especially when they're making a decision about what will happen to a particular student.
- Colleges using data analytics have to make sure their students have "open futures" - that their programs create educational opportunities, not the other way around.
Colleges aren't just using big data to scope out which students are likely to struggle and which are likely to succeed. They're also crunching numbers to track attendance, evaluate their curricula, identify likely donors and investors, and notify students who aren't aware they're able to graduate.Even admissions offices are finding ways to incorporate data analysis into their programs, using everything from test scores to social media to determine which applicants are most likely to matriculate and which matriculants are likely to stay.Advertisement
- Assam making all-out efforts to provide best healthcare services to COVID-19 patients: Sonowal
- NEET postponed to Sept 13 in view of COVID-19 , JEE-Mains to be held from Sept 1-6: HRD Ministry
- Record spike in daily COVID-19 cases in Nepal; total crosses 15,000-mark
- Odisha fixes Rs 2,200 for COVID-19 test in private labs
- COVID-19: Flying squad to keep check on Bhiwandi hospitals