Jon Reid acknowledges he's a bit of a procrastinator, and his study habits can be defined as "more loose and less structured." So when the now-third-year senior learned during his freshman year about a personalized education tool being offered at his school, the University of Michigan (U-M), Reid was immediately on board.
The second semester of Reid's freshman year, he took a Statistics 250-level course, which he says is "notoriously difficult at Michigan." Although Reid is a history major, he was required to take a quantitative reasoning class. "The first day I went into a lecture hall of maybe 300 students and I felt completely overwhelmed," he recalls. "Math is not my strong suit."
Reid says the professor showed a quick tutorial on ECoach in the lecture hall, and students were even offered "a very small amount of extra credit to use it." The concept of receiving personalized support with resources and a checklist of what to do before an exam was appealing because it's "not my personality to go up to a professor."
ECoach helped Reid keep "everything fresh and centralized" and created a plan of attack for the class. "I felt I was staying ahead, and I was very confident going into the first exam." The tool even sent Reid personalized feedback on how he did (he received an A-on the first exam), including the median grade, where he fell and some tips to improve his score.
"I was very taken aback by that. I have never had a class where there was follow up on an exam with feedback and encouragement," he says.
ECoach is the brainchild of U-M professor Timothy McKay, who began looking at student data in his large Introduction to Physics classes in 2008, trying to understand who was succeeding and who was struggling. "Looking at the data made me recognize the differences in backgrounds and goals, and the reasons for taking physics and the affect toward it," McKay says. Some students were enthusiastic and some were terrified, he says.
"I found myself wanting to be speaking differently to every one of my students," McKay says, "and do that in a way that was informed by who they are, where they're coming from and their trajectory; all things you'd like to know in order to coach them effectively."
ECoach is designed for first-year students taking science, technology, engineering, or mathematics (STEM) classes, who can access the ECoach website with a single sign-on. The personalization ECoach offers is based on data the university already has, such as what courses a student has taken, what their grades have been, and information from their admissions application, "so we can note what their high school background was and their standardized testing" scores, according to McKay.
Students are also asked questions, such as whether they are frightened about taking a particular class. If so, "it's important that we talk to them and let them know if they work with us and follow the advice [in ECoach], they can be successful," McKay says.
Today, it is becoming the norm rather than the exception for colleges and universities to utilize the data they have within student learning management systems (LMSs) and student information systems (SISs) for academic purposes. The reasons are not entirely altruistic; yes, higher education officials want to help students be successful while in school, but they also want to do whatever they can to keep them there, so the dollars continue to flow in both from students and from state and federal funding.
Data analytics is being done more frequently because "people are starting to really get a sense of how urgent it is" to keep students in school, says Glenda Morgan, a senior director and analyst in market research firm Gartner's Higher Education division.
"Tuition is a big income stream and the proportion of state funding has gone down dramatically and the portion students are paying has gone up," she says. There has also been a mind shift from not just getting students into college but "increasingly, it's 'Okay, we got them in the door; let's get them to succeed and get out on the other side,'" Morgan says.
Also, the analytics tools have gotten better, so campus officials are becoming more knowledgeable about what it takes to help students, especially those at risk of dropping out.
Although analytics in higher education is still in the early stages of implementation, 40% of CIOs say they will receive increased funding for it in 2019, according to Gartner research on the "Top 10 Business Trends Impacting Higher Education in 2019."
"More and more demand is being put on campuses to use analytics to improve student success ... because higher ed campuses are responsible or accountable to numbers of entities" such as accreditors, the federal government for financial aid, and their state for funding, notes Linda Baer, a senior consultant in higher education.
Some states even use performance-based funding for higher education, says Baer. Another compelling reason: in addition to stakeholder accountability, the traditional student population is declining, she says. "One way to counter the [decline in] incoming enrollment is to improve retention and graduation rates, so people are focused on 'how do we keep students staying here instead of dropping out?'"
Right now, one of the hottest uses of data is nudge technology, which incorporates behavioral economics principles to help enhance student outcomes, Morgan says.
Student success was listed as one of Educause's top 10 most pressing IT issues for 2019, according to Kathe Pelletier, director of Student Success Community Programs at the non-profit higher ed association.
"We interpret this as institutions taking responsibility for students' success, not just in the classroom, but as a whole person," Pelletier says. "Responsible use of trusted data is a key enabler here."
In terms of deployment of analytics technologies, only 7% of institutions have deployed predictive analytics for student success institutionwide and 58% are expanding or planning to do so, according to Educause's 2019 Strategic Technologies report, Pelletier says. Course-level learning analytics are slightly less widespread, with 3% institutionwide deployment and 35% planning or expanding, she adds.
"From these data we might expect student success analytics to become widespread at some point, but the additional complexity of measuring learning, potentially requiring faculty to approach assessment differently, may slow down the deployment of course-level learning analytics," Pelletier says.
Universities are certainly feeling the pressure to help students succeed. In a 2018 joint study by the National Association of Student Personnel Administrators, the Association for Institutional Research, and Educause, 88% of 970 senior-level student affairs professionals strongly agreed or agreed that to stay competitive, they must continue to invest in student success analytics. Among the respondent institutions' goals for conducting student success studies, 96% said they wanted to improve student outcomes from interventions, 71% said they sought more efficient delivery of programs or services, and 39% cited elimination/reduction of programs shown not to contribute significantly to student success.
At the University of San Francisco (USF), student data within the LMS such as attendance, grades, and assignments are being used to determine their progress, says vice president and Chief Information Officer Opinder Bawa. "For the first couple of weeks, we're very vigilant on [capturing] that information so we can do early intervention," he says. If a faculty member notices a student struggling, they can use technology developed by USF's IT department to alert school counselors to reach out to that student, Bawa says.
USF IT also built a mobile app called USFMobile, which pulls data from USF's Salesforce system used for advising and counseling, and can generate alerts, he says.
The only people who have access to USF's LMS data are faculty members, and counselors have access to Sales-force. The university is working on putting in place rules for data governance, Bawa adds.
Eventually, student data also will be used to "really understand what [students) are doing, and what we can do to help them: Are they involved in clubs and sports?" he says. "If not, how can we get them involved and to be more successful at the university?"
However, Bawa adds, "We have to be very thoughtful in which data to use and how much data to use, so we don't cross the line into privacy."
Right now, USF has an 84% graduation rate in six years; Bawa says the goal is to "move the line" higher.
The University of Missouri is piloting a 12-question student success survey this year to help new students integrate to the campus. The purpose is to glean how students are adjusting in terms of academics, social, and financial concerns, and whether they are satisfied with the institution and plan to return, says director of Strategic Initiatives and Assessment for Student Affairs Ashli Grabau,. The data is stored in the university's SIS.
If a student indicates any concerns in the survey, the information is used "to reach out to them to make a high-touch response," says Grabau. The impetus for the survey was "to really improve our student success, especially for new students, and for them to feel connected to the institution and get the resources they need to feel financially secure, and resources to improve their academic experience and improve their sense of belonging," she says.
The top three issues the survey revealed were focused around course struggles, class attendance, and financial concerns, according to Grabau; there were no surprise findings. "I think it reinforced that it's important to check in with students early on to see what they're struggling with and intervene, rather than halfway through the semester."
Although the university had a record 87.9% students return this year from last year's freshman class, "It's not as high as we want it to be," Grabau says. "Our goal is 93% and we're doing pretty good, but to move from good to great takes ... a high-touch approach."
Higher education observers say that as student data is more frequently being sliced and diced, it raises concerns about privacy, interpretation, and misuse.
"One reason students aren't successful is because they take courses that aren't useful or they change their mind about their major," says Gartner's Morgan. Some predictive analytics tools suggest courses of study and majors based on profiles of others who have been successful in those majors, she says. "They can be hugely successful, but I do worry that sometimes people aren't questioning the assumptions behind algorithms."
One of the biggest potential risks with student data is related to equity, says Educause's Pelletier. "Disaggregating data allows institutions to identify and address potentially hidden opportunity gaps for historically underrepresented students. When institutions are not appropriately disaggregating data based on race, ethnicity, gender, and socioeconomic status, not only do they risk missing insights about the needs of various subpopulations, but they risk furthering the structural barriers and increasing equity gaps."
Another equity concern related to data is in the way institution officials talk to students about what they "see" in their data, she adds. Consequently, adequate training for advisors, faculty, and other support staff in how to provide culturally responsive guidance is critical to promote a sense of belonging for students, which is a key driver of retention, Pelletier says.
Additionally, "Sharing the wrong data with students, or sharing the right data in the wrong way, can lead to alienating experiences," Pelletier says.
While Baer believes in "know your students, know your data," she says the next big principle is having data governance and policies in place. "That is how you secure the data, because a big tenet is to keep student data private," she says. "And it's the law."
"Analytics systems are the new, bright shiny object of higher education, but it's a new way of looking at the data," and officials must maintain that tenet as the demand grows for student information in real time, agrees Colleen Carmean, founder and president of the Ethical Analytics Group, an organization created to help higher education create student and institutional success via data. Since universities often do not have staff with experience or thoughtfulness on how to use data ethically, a new trend is the rise of the data privacy officer, says Carmean, who is also former associate vice chancellor for academic innovation at the University of Washington-Tacoma.
"Disaggregating data allows institutions to identify and address potentially hidden opportunity gaps for historically underrepresented students."
Carmean recommends universities make certain students are aware of the intentions of any data initiatives being done on their behalf, and to be clear about motivations when sending out email messages asking for information.
"We may think we know what's good for them, but it's their data," Carmean says.
By and large, universities are on top of protecting student data, Baer says. "Our trouble is that computer systems are as strong as you can make them, but there's always security problems so campuses have to stay very vigilant on that."
Many campuses also have ethics committees, so there are policies for how student data can be used, as well as who can use it, for what purpose, and how they protect it, Baer says.
Data management and governance was ranked #8 on the Educause top 10 issues list in 2019. Safeguarding student data has been a theme for a while, but the type of data available and the number of systems using and storing data have increased in volume and complexity, Pelletier says.
There are resources that can help. For institutions using cloud services, for example, the Higher Education Cloud Vendor Assessment Tool (HECVAT) can provide guidance in managing risks to the confidentiality, integrity, and availability of sensitive institutional information and the personally identifiable information of constituents, according to Pelletier.
Ultimately, as schools migrate toward a student-focused experience and student success initiatives continue to gain momentum, they will continue seeking ways to engage students in defining what success looks like for them. Schools will also continue leveraging technology to send nudges, reminders, and resources based on each student's own goals, Pelletier says, all in the name of helping them feel more connected.
Citing the quote "With great power comes great responsibility," she says the promise of using trusted data to drive outcomes in student experience and student success must be tempered with employing an ethical approach to the collection, storage, and use of data.
"Privacy and security are important for all," Pelletier says, "yet higher education has a particular responsibility to ensure our data practices are making institutions more student-ready, especially for historically underserved students."
Baer, L.L. and Carmean, C.
An Analytics Handbook. Moving From Evidence to Impact, 2019, The Society for College and University Planning. http://bit.ly/2Cfyl7R
Responsible Use of Student Data in Higher Education, Stanford University CAROL & Ithaka S-R Project, 2018. http://ru.stanford.edu/
Ekowo, M. and Palmer, I.
The Promise and Peril of Predictive Analytics in Higher Education, 2016, New America. http://bit.ly/2pqwxpx
Defining Student Success Data. Recommendations for Changing the Conversation, 2018. Higher Learning Commission. http://bit.ly/34v7i4s
Institutions' Use of Data And Analytics for Student Success, 2018, National Association of Student Personnel Administrators, the Association for Institutional Research, Educause. http://bit.ly/36DiWMm
©2020 ACM 0001-0782/20/4
Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and full citation on the first page. Copyright for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or fee. Request permission to publish from email@example.com or fax (212) 869-0481.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2020 ACM, Inc.
No entries found