Page 31 - Campus Technology, March/April 2019

and Planning administers two surveys, one on the impact of orientation and the other to baseline student engagement or their pre- paredness for college. Then there’s the first- year experience survey given during the fall semester as well as other surveys on financial literacy, student satisfaction with services, satisfaction with admissions and enrollment for transfer students and plenty more. Plus, every semester students’ GPAs are updated, grades are added, data on math proficiency is factored in, and scores from the Multi-State Collaborative to Advance Quality Student Learning are collected, along with standing on the dean’s list and majors declared.
Each student is tracked for at least seven years, including where they’ve gone if they’ve withdrawn from the institution before gradua- tion — whether it’s to a private in-state uni- versity, a community college, a college in
some other part of the country or out of high- er education altogether, according to Michael Ben-Avie, former head of the Office of Assessment and Planning (Ben-Avie is now senior director of learning assessment and research at Quinnipiac University).
The cohort data is uploaded to IBM’s Watson Analytics, which suggests starting points for data analysis, such as “What drives retention.” When a starting point is chosen by the user, Watson will produce a visualization with the most important predictors and give each one a “predictive strength.” From there, the user can ask follow-up questions using natural language. Ben-Avie tells students in the school’s Watson Analytics intern- ship program, “You don’t need to know statistics because Watson will pick the most appropriate statistical analysis to run. You don’t need to code because it’s natural language. What it does is get at broad strokes,” he said. “Then I’ll go to use SPSS and other statistical programs to probe and explore more in depth.”
Armed with a data-derived understanding of what the university needs to pay attention to, the school engages in “campaigns,” programs to
address specific findings. For example, take a student who has expressed her intent to go into nursing, but has struggled with the prerequisite courses. The campaign might be, “You’ve always wanted to go into nursing, but [why don’t you] explore other careers in health fields?” As Ben- Avie noted, “We don’t have to say to the stu- dent why [she] received that particular commu- nication. But we know the reason is because that student is not likely to be accepted into the program and we don’t want that student to leave higher education.”
Another campaign might target first-generation college students, who are, said Ben-Avie, “vul- nerable in terms of withdrawal from higher education. They may not have the support of their families or their immediate environment.” The university has created “living-learning” communities where students live in the dorm with residential staff members who were first- generation college students themselves. Stu- dents also attend courses with professors who were as well. “We don’t say, ‘The reason why we’re [encouraging] you to [join] this living- learning community is because you’re vulnera- ble,’” he emphasized. “We approach them and say, ‘We have this really great opportunity and we think it’d be great for you and you would gain so much from it.’”
And Ben-Avie pointed out that the living- learning communities aren’t a foregone recom- mendation for every first-generation student — just the ones for whom they’re best suited.
Category: Administration
Institution: Southern Connecticut State
Project: Educational Intelligence: How One University Uses Data to Improve Student Outcomes
Project lead: Michael Ben-Avie, former director, Office of Assessment and Planning
Tech lineup: Digital Measures, IBM, Snap Surveys, Watermark
University
Photos: Southern Connecticut University
campustechnology.com 31