If data about struggling students is to be used in a way that supports their mental wellbeing rather than harms it, what kind of data do learners want to see and what actions do they want it to trigger? We find out from projects that have talked to their students to discover just that.
Student mental wellbeing is an increasing concern for universities. A 2015 National Union of Students (NUS) survey1 found that eight out of ten students (78%) said they experienced mental health issues in the last year. A third said they would not know where to ask for mental health support at their college or university if they needed it, with 40% reporting feeling nervous about the support they would receive from their institution.
Poor mental health has an impact on every aspect of a student’s life, from feelings of social isolation to academic failure. According to the Higher Education Statistics Agency (HESA), 1,180 students who experienced mental health problems left university early in 2014-15, the most recent year in which data was available. That’s a 210% increase from 380 in 2009-10.
However, students tend not to ‘drop out’ of university without any warning. The final decision to leave is likely to be preceded by a period of dwindling attendance, late submission of work and falls in grades. If these signs were picked up sooner by personal tutors, or another member of university staff, there’s a chance that at-risk students could be better supported. Could learning analytics help to alert tutors earlier to struggling and vulnerable students?
Samantha Ahern, learning technology project officer at University College London (UCL) believes so.
“Learning analytics can be used to help to address some of these inadequacies by providing timely and meaningful data to personal tutors about their tutees [which] is in alignment with the Universities UK guidance…to align learning analytics with student wellbeing,”
she says. However, she warns,
“there are legal questions still to be answered around negligence and failing to act on or engage with information provided via learning analytics.”2
Maintaining an ethical approach
Julia Taylor, Jisc subject specialist in accessibility and inclusion, highlights that Jisc has a learning analytics code of practice, providing an ethical approach to gathering student data, and suggests some of the questions that might enable the right metrics to give early warning of mental health concerns, allowing timely responses.
For example, has a student stopped attending? Or stopped using the VLE? Is there a declared disability or a known risk? Are there factors in course design or content that created stress and contributed to the change in behaviour? How best should it be responded to? How do known patterns of engagement correlate with wellbeing?”3
Interpreting the data fairly
However, care must be taken that neither the data use nor any interventions put in place as a result of it exacerbate the issue. Some researchers have warned that misguided use of comparisons in learning analytics dashboards could be an additional source of stress for vulnerable students.
It’s a topic that the University of Huddersfield’s director of learning and teaching Liz Bennett has been exploring. In a small-scale, qualitative study with students from the university’s School of Education, she focused on the design of dashboards and how students understood and interpreted the data that they presented.4
“We know there can be a huge amount of emotional response to getting feedback on assessments and I was interested to see how they responded when it was in a dashboard format,”
What was unusual about the dashboards in the study was that they presented the data – third year students’ grades across their degree – in a comparative format, showing each student how they performed in relation to the rest of their cohort.
Surprisingly, perhaps, although the impact of seeing grade feedback was indeed emotionally charged for some students (“The saddest one is the core summary overall because looking back on grades that you’ve previously had - you can’t really change them any more so you can’t really do anything,” says Ingrid, who came 168th out of 178) the overwhelming impact was motivational, particularly for the lower performing students.
“You might think that if you came first out of 178 you’d be pleased, which indeed that student was, but it was also true that when people were coming near the bottom they weren’t pleased, necessarily, but they were motivated by seeing it and it gave them an insight,”
says Dr Bennett.
“There were some students who didn’t want to see the comparison but there was also evidence that it was motivating.” As one student, Marcia, who came 53rd out of 178, says: “I think as soon as I saw it I decided I’m taking a month off [paid] work to just get on with my dissertation”.
Getting students involved
As a result of the study and the feedback from students Dr Bennett strongly believes that such dashboards need to be customisable by the student so that they can choose whether to compare themselves to the whole cohort or the top 10% or the bottom 10% – or to see no comparison data at all – to avoid any negative impact on a student's wellbeing or reinforcing any feelings of negativity.
“I think we need to give students choice about what they see to make sure those who are vulnerable have got some control over whether they see themselves compared to other people or not. There is potential for it to go wrong because you are dealing with emotionally charged information so it does need to be scaffolded and supported in the way it’s rolled out,”
advises Dr Bennett. She also notes that, in her study, the students’ reactions were gathered through face-to-face interviews, and further research is needed on what might happen if students were to get such dashboards unmediated, without the opportunity to talk through their emotions about them immediately.
How and when students might want to talk through issues relating to their studies has interested Sarah Parkes, tutor for transition and retention/foundation year tutor at Newman University Birmingham. Working through the university’s student-staff partnership framework with three 'Students as Partners' projects, she’s been surprised to discover that, overwhelmingly, what students want is tutor and peer proactive mentoring systems that respond when data is suggesting that someone is falling behind or in need of support.
“Peer mentoring came out as a key intervention that students thought would be valuable. They would be happy to have other students get in touch to support them,”
“Coming to university is quite a big step for a lot of students and so I think they feel better about the idea of us using other students to support students as they felt like they wouldn’t be made to look silly – if it was a member of staff it may have felt punitive but if it’s another student they felt more comfortable talking about things.
Our focus on analytics has to sit with our ethos about being student centred with the person at the centre. Whatever we do in terms of an intervention needs to be supportive – and part of a wider mechanism for support – rather than potentially feeling like it was punitive.”
Students also had strong feelings about who they wanted to contact them if the data flagged up that there might be an issue. Any communication triggered by the data had to come from someone they had heard of – a member of staff or student from within their own department with whom they might have had previous interaction – rather than at the broader institutional level, such as student support or registry.
While students were, perhaps surprisingly, mostly relaxed about the use of their data – if the right people were getting it and using it appropriately then they were, in principle, happy – Sarah Parkes is keen to stress the sheer complexity of this kind of work in terms of achieving a holistic sense, through the data, of who the student is and what they are doing and where collaborations are needed. However, the benefits – for students, for tutors and for the wider mental wellbeing agenda – are undeniable.
“Talk to your students! That’s been a real test of what we’ve done and helped us to understand how our students feel about this work,”
urges Sarah Parkes.
“Don’t try to deal with it in isolation.”
- 1 The NUS surveyed 1,093 students in further and higher education in November and December 2015 on behalf of the All Party Parliamentary Group (APPG) on students - https://www.nusconnect.org.uk/resources/mental-health-poll-2015
- 2 UCL Digital Education Team blog: Learning analytics as a tool for supporting student wellbeing - https://blogs.ucl.ac.uk/digital-education/2017/11/20/learning-analytics-...
- 3 Jisc accessible organisations blog: student minds, how can tech contribute to better mental health? - https://accessibility.jiscinvolve.org/wp/2017/09/25/mentalhealthtech/
- 4 Read Liz Bennett’s report, students’ learning responses to receiving dashboard data - https://www.srhe.ac.uk/research/srhe_funded_projects.asp