What can we learn about student mental health from surveys?

​Universities are starting to run surveys to monitor the wellbeing of their student populations. These are powerful resources that will enable us, in time, to test whether student mental health is changing. For example, whether a current cohort of students are more likely than previous ones at a particular institution to be anxious or depressed. The data could also be used to track student trajectories as they progress through their university years, allowing interventions to be targeted at times and groups of greatest need.

What surveys of students won’t be able to tell us is whether students are different from the rest of the population. If the mental health of those in higher education is deteriorating, does that just reflect wider trends in the population? Or might changes over time in the mental health of the student population be explained by changes in the characteristics of who becomes a student?
 
One way to explore whether students are different is to compare them to people of similar age who are not students. The UK has a rich and long tradition of national general population surveys, with many datasets freely and instantly downloadable from the UK Data Service’s central repository. It’s possible that among these there are some that could be used to compare students and non-students, although we’re not aware of that having been done before.
 
I’m based at the National Centre for Social Research, and we run many of these large-scale data collections for various Government Departments and agencies. We try to anticipate secondary data needs when designing questionnaires – treading a difficult line between maintaining comparability with how questions were asked before (so change over time can be traced) with covering topics in ways that are meaningful in the current context.
 
What we hadn’t anticipated till recently when designing these surveys was quite how significant a group higher education students were about to become, both in terms of national policy and public interest. So a key limitation is that students are not always identifiable in the sample. For this reason we have prepared a resource which sets out the existing surveys we are aware of that do allow students and non-students to be compared. These data sets do come with limitations, and these are also set out in the document. 
 
So far we have found little evidence of general population surveys being used to examine the circumstances of students. This must have been done – and we’re keen to hear from you if you know of examples.

​If you’re interested in exploring how we can best make use of general population surveys to understand what is distinctive about student mental health, join us for a webinar at 11am on 21 March. Registration details can be found here.


Sally McManus is an Associate on the Health Team at NatCen Social Research. She is interested in the measurement of mental health and wellbeing.