Student wellbeing is a complex, multi-faceted concept defined in different ways across jurisdictions and sectors. Not only does the lack of a commonly agreed definition hamper policy coherence, it impedes accountability, and efforts to monitor and evaluate policies, programs and outcomes. Conceptual frameworks that define cross-sector wellbeing indicators do exist, such as the Australian Research Alliance for Children and Youth’s (ARACY) ‘The Nest’ (presented below) and the UNICEF Innocenti wellbeing framework.  However, such frameworks are yet to be adopted on a cross-sector basis in Australia.

Source: Adapted from ARACY’s ‘The Nest’ 2014

The determination and adoption of a national definition of student wellbeing is necessary not just to elevate its importance across jurisdictions but to enable better cross-sector collaboration. A clear and shared understanding of what student wellbeing means is imperative to avoiding varied interpretations and to fashioning, implementing and monitoring effective policy responses.

No single indicator of a complex concept like student wellbeing will ever be sufficient. So, it is important for policymakers to consider a suite of quantitative and qualitative indicators in order to reliably inform decisions.

Education systems often rely on reactive or lagging indicators of student wellbeing as they tend to be more readily observable and measurable. Examples include prevalence of wellbeing-related incidents reported in schools, rates of mental health issues, or usage of support services. Such indicators, while instructive, are of limited use to decision makers committed to proactive and preventative strategies.

By contrast, leading indicators are predictive of outcomes that are likely to happen in the future. Capturing the presence or absence of well-evidenced risk and protective factors can help reveal the potential need for support. For instance, high device screen time, adverse life events (such as exposure to violence), poor family functioning, and high demand academic environments have all been found to be predictive of diminished student wellbeing.

Ultimately, both lagging and leading indicators are necessary to capture the breadth of evidence required to inform wellbeing responses in schools. It is through recognising the strengths and role of each – and utilising them accordingly – that evidence-driven improvement will emerge.

Validity is also an important consideration – that is, whether an indicator measures what it is intended to measure. Validity is not a feature that exists on its own. Rather, it is defined with respect to the question(s) asked.

For instance, the number of unexplained days of absence at school is a perfectly valid measure of student attendance. However, if the question is about student engagement, an indicator based on absences has reduced validity, as students may have perfect attendance, yet be disengaged. Extending this example to wellbeing decreases the validity even further, as students with zero absences and high engagement may struggle with wellbeing.

The concept of validity is also key in the context of cultural appropriateness. In Australia, it is well evidenced that cultural connectedness is a significant wellbeing dimension for Aboriginal and Torres Strait Islander communities. This raises questions about the validity of any set of wellbeing indicators applied in Indigenous contexts that are based on dominant Western frameworks.

Assessing student wellbeing is ultimately a sensitive and often subjective matter. Strong multi-disciplinary relationships and in-depth understanding of the student context are critical to delivering effective, well-targeted responses. If supported by appropriate data infrastructure (both from the perspective of usability and data privacy), rich place-based qualitative data can be collected in a systematic way that can inform key decision-makers at a system-level.

Attribution – that is, ascribing causal relationships between variables – is at the centre of almost every policy discussion seeking to understand the effects of a given policy or program.

While determining attribution can often be challenging, it is especially so in the context of wellbeing. Attribution of changes in student wellbeing to a specific educational intervention is made particularly challenging by how dependent student wellbeing is on environmental factors. A recent Report Card by UNICEF Innocenti highlights the importance of viewing a student’s wellbeing in the context of their ‘worlds of influence’. Student wellbeing at school cannot be viewed in isolation from a broader context of their family, community, and society. Without a way of accounting for these factors, drawing conclusions about causal relationships between policy initiatives and student wellbeing is extremely difficult.

It is also important to critically reflect on potential causal direction. Reflecting on the link between preventative services and referrals is a clear illustration of this problem. Student referrals datasets are one of the richest sources of information on the frequency and type of support required by students. It is easy to jump to the conclusion that better prevention should cause a reduction in referrals. However, prevention activities might also result in greater awareness about the supports available and therefore higher referrals.

Quality data as an enabler to system-wide change

While the last decade has seen significant advances in the collection and use of evidence to drive performance improvement across education systems, the collection, interpretation and use of evidence in relation to wellbeing remains an area where further progress is needed. The imperative to achieve this progress has been accentuated by the events of the last year.

Ultimately, the success of our education systems to, in Andy Hargreaves’s words, “move beyond an age defined by achievement and effort alone, to one that also gets to grips with engagement, identity and wellbeing” will largely be underpinned by a genuine commitment to the development and effective use of a coherent and robust evidence base.

This article was authored by Pola Orlowska and Rachael Mariani from Deloitte Access Economics.