Colleges have the ability to gather more student data than ever before, and with that information comes the pressure to put it to good use. Nudges, when done right, are proving up to that task.
Nudging is a concept that has gained traction in higher ed ever since the 2008 release of the book "Nudge: Improving Decisions About Health, Wealth, and Happiness" brought the subject to the public's attention.
Put simply, nudges are interventions that steer someone toward a better decision without taking away their choice. In higher ed, they take the form of messages delivered through texts, emails or the learning management system that warn a student if they've fallen off track, alert them to important deadlines and make them aware of campus resources. Student data comes into play by helping colleges figure out who needs what kind of information most.
College leaders are increasingly looking to nudges as a way to buoy student success. So far, the results are promising, with various studies finding they've helped to reduce summer melt, boost enrollment and increase retention. But experts on nudging warn it isn't a magic bullet — and can even push students out of college if the nudges aren't well-designed.
"We have to make sure that all this work is designed in a way where it's 'do no harm,'" said Mark Milliron, the co-founder of Civitas Learning. "This is really about helping that student make a good choice for them."
Good nudges get good results
When nudges are targeted and executed properly, they can help colleges achieve progress on some their most hard-to-solve issues.
Take Winston-Salem State University, an HBCU in North Carolina that uses AdmitHub’s AI chatbot to send incoming students text messages about the tasks they need to complete before they step foot on campus. It also provides current students with information about the university's resources.
Joel Lee, the university's assistant vice chancellor for enrollment management, credits AdmitHub with helping find a solution to one of the university's pressing problems: students showing up to campus without having completed basic enrollment requirements.
As a result, Lee said, those students "were spending that first week of school ... trying to take care of basics rather than focused on the new student experience and getting to know the campus and making friends."
The university's chatbot is part of its larger Ram Ready initiative, which aims to help students complete three key steps before they can move into their dorm: pay their bill, ensure they have the required immunizations and submit their final high school transcript.
The chatbot acts as a virtual assistant to help answer students' questions about how to fulfill those requirements, pulling answers from its information database or passing along those questions to a university staff member. So far, the chatbot can answer roughly three-fourths of students' questions on its own, Lee said.
Readily accessible information is of particular importance at an HBCU such as Winston-Salem State where a large share of incoming freshman are first-generation college students who may not have someone at home who can guide them through the enrollment process.
In the fall of 2018, following the university's Ram Ready and chatbot rollout, college officials saw a 74% year-over-year spike in the number of new freshman who had fully met their financial requirements, along with a 37% increase in immunization compliance.
After nudges help get students through the door, they can help them stay there.
The University of Washington Tacoma partnered with Persistence Plus to develop nudges to help first-year students stay on track. They are piloting a similar effort for students who are a few credits shy of being able to graduate.
So far, the university has used nudges that have ranged from reminding students about Pell Grant deadlines to notifying military veterans on campus about free donuts, simple gestures a college can use to foster a "sense of belonging," said Colleen Carmean, the university's associate vice chancellor for academic innovation.
The effort has worked. A 2017 study on students close to graduating found colleges using Persistence Plus's text message-based nudges — including U of Washington Tacoma — saw a 6 percentage point increase in degree completion among students most at risk of dropping out.
Changes in who is attending college are one driver of the growing use of nudges. Once reserved solely for the elite, college is now marketed as a pathway for low-income and other underserved populations to secure a high-paying job and a better future.
But colleges have been slow to respond to the different needs of this new generation of students.
"Until recently, most universities have taken the attitude that if a student doesn't succeed, it's the student's problem when, in fact, we're taking their money and we're letting them down," Carmean said. "There are some universities that are doing a really good job of understanding that we have some responsibility to know our students, to offer services that will assist them in making the right choices and to get them to the finish line."
What makes a good nudge?
A college's responsibilities don't end with setting up support services. They also must make sure those services aren't backfiring.
For instance, students who are barraged with daily messages that don't pertain to them may opt out of the service, limiting a college's ability to reach them. "You can end up with nudges turning into noise," said Civitas' Milliron. "And when nudges turn into noise, there's no impact."
Even personalized messages can push a student away. For example, sending students an email alert that they will be dropped from a course if they miss one more class could drive them to stop going.
"A student ... whose kids got the flu and they've missed a week and a half, they get that message and they go, 'Oh God, it's just not worth it,'" Milliron said. "They just quit."
To design effective nudges, colleges should start by identifying which populations face the biggest roadblocks to their success.
With troves of student data, it's now easier to see which groups are showing up to class and which aren't, among other measures of student success. What's harder — but usually far more useful — is finding out why.
As Milliron noted, students may not miss class because they are forgetful but rather because they are juggling life's responsibilities or dealing with emergencies. In those cases, nudges that simply remind students of class times and policies won't do.
"Until recently, most universities have taken the attitude that if a student doesn't succeed, it's the student's problem when, in fact, we're taking their money and we're letting them down."
Associate vice chancellor for academic innovation, University of Washington Tacoma
An example outlined in the Society for College and University Planning's "An Analytics Handbook: Moving from Evidence to Impact," which Carmean co-edited, recalls how one academic advisor was able to help a student obtain emergency funds to replace her car tires only after she responded to a follow-up nudge about having transportation issues.
Similarly, a student struggling in a class may not benefit from a nudge listing the time and location of the tutoring center if they feel as though it would be admitting failure to show up there. A better nudge may try to develop positive social norms around that targeted behavior, such as by noting that students who go to the tutoring center tend to get the best grades, Milliron said.
It's no small feat to use the data to figure out where nudges are needed and if they work. And if institutions don't have the capacity to monitor that data, they should forgo nudges until they can, experts said.
At U of Washington Tacoma, a team representing various corners of the administration met about once a week to do just that. "The more people you have at the table, the more ideas you have for just who could we pull up next and see how they're doing and what could we do to nudge them to take some action in their own interest," Carmean said.
However, Milliron cautions colleges against viewing even the most well-designed nudges as a cure-all for the broader woes hampering student success.
"Nudges aren't going to save the world," he said. "Nudges are a strategy in a larger family of activities. It's the right kind of information. It's the right kind of support systems. It's the right kind of curricular resources ... combined with the right nudge at the right time."
'It's a double-edged sword'
In the same way nudges can target behavior beneficial to the student, they can be turned around to favor only the institution.
In 2016, an email exchange revealed a proposal by Simon Newman, then the president of Mount St. Mary's University, to use student data to dismiss those deemed unlikely to succeed, The Washington Post reported. If the university could get between 20 and 25 students to drop out before late September, the deadline for submitting enrollment data to the federal government, it could increase its retention rate by roughly 5%, Newman wrote to faculty, according to the student newspaper. He later resigned amid the controversy.
Good nudges should "improve the welfare of those being nudged, not those doing the nudging," contended New America's Ernest Ezeugo and Iris Palmer in an article published in EdSurge last year.
Though the Mount St. Mary's example is extreme, colleges can end up using nudges unethically whether they intend to or not. And those cases may fly under the radar because institutions don't want to advertise their failures.
"You can end up with nudges turning into noise. And when nudges turn into noise, there's no impact."
Co-founder, Civitas Learning
As Ezeugo and Palmer note, an ethical nudge gives students choices instead of directives, is transparent and comes with extra support.
Bias is also important to watch for, Ezeugo told Education Dive. Colleges should train staff to know the data's limits and to avoid having their implicit biases affect what nudges are sent to what groups.
The pressure to compete over a shrinking population of undergraduates doesn't help, Ezeugo added. In a dash to improve their student outcomes data, colleges may view nudges as a quick and low-cost solution. But that can lead to trouble.
To be effective, Ezeugo said, nudging requires constant testing and oversight from a team whose members can bring different perspectives to the table.
"It's a double-edged sword," he said. "It really does take careful consideration, careful thinking to deploy nudges ... in a way that doesn't harm students."