‘Warm data’ is a concept developed by Nora Bateson in an attempt to improve the ways in which conventional forms of statistical data can be used to address issues within complex social systems. The term is a response to the rise of ‘Big Data’ and the increasing credence now being given to analysis of aggregated data that has been taken out of context. In contrast, ‘warm data’ is other information that needs to accompany any bald statistical analysis.
As such, ‘warm data’ reminds us that analysis of statistical data taken out of context can never fully correspond with the complexity of any ‘living system’ – from an individual school through to our whole-planet ecology. Thus, the idea of warm data raises the idea of complementing decontextualized official ‘data’ with a nest of other information that can gives a sense of how any system is functioning in terms of its broader relationships and interdependencies.
Bateson points idea that these issues of relationality are rarely talked about in terms of data analysis. As such, the idea of ‘warm data’ is designed to draw attention to (and make us comfortable with) the idea of other forms of information that can augment and work with existing forms of decontextualized official ‘data’. As Bateson puts it, any presentation of statistical data should always be accompanied by the question of ‘… and what is the warm data on this?’.
So, in terms of our own research, ‘warm data’ reminds us of the need to strive for uses of data that are better grounded in the interdependencies and inter-relationships that exist within and without any school, and that ultimately work to fully reflect the school as a living system.
For example, even something as commonplace as school data relating to ‘learning’ can be seen as a complex (and partially unknowable) issue that requires nesting within ‘warm data’. At the moment, generating information on ‘learning’ is largely reliant on the official statistical measurement of exam, test and assessment results. Some schools might also include evidence of engagement with learning through the official ‘learning analytics’ functions of educational software, or other ‘trace data’.
Yet we also need information on the relationships between the many different contexts that impacted on students’ ‘learning’ and how these statistical data came to be. These are likely to include family relationships, peer relationships and other relations across a school community. These are also likely to include biological and ecological inter-dependencies. These are also likely to include political, economic and cultural relationships at a local, regional, state, national and perhaps even international levels.
From neurodiversity to institutional racism, there is a lot that needs to be ‘taken into account’ when seeking to generate information on students’ learning – different kinds of information that give context (or multiple contexts) to the statistical data that are routinely produced and used within a school.
This is currently being foregrounded for the present ‘Class of 2020’, whose learning is being profoundly impacted by the impact of the summer bushfire disasters followed by the on-going COVID-19 disruptions and associated traumas. As is beginning to be acknowledged by education systems around the world, any ‘learning’ data and official performance measures generated from the current period of the pandemic deserves a lot of contextualisation.
It is important to note that this approach will not result in clearer analyses, concise answers or neat facts. This approach will often highlight contradictions, gaps and point to what we don’t know. Nevertheless, the idea of ‘warm data’ acts as a prompt to think differently – and more expansively – about the information that can be used within a school.
The exact ways that schools can make use of warm data need to be worked out on a ‘bottom-up’ basis. Schools need to first recognise that this is a realm of information that is important, and then work out ways that this sort of information can be made.
Methods-wise, the idea of ‘warm data’ fits well with our project’s interest in participatory design. Bateson advocates for the running ‘Warm Data Labs’ where various stakeholders can come together to mutually discover and co-construct forms of warm data relevant to school life.
Theoretically, this approach is borne from a cultural anthropological basis (Nora Bateson runs an ‘International Bateson Institute’ named after her father and renown anthropologist Gregory Bateson). There are a number of cognate approaches here – including
- Multiple description
- Iterative multi-modal learning
- Ecology of communication
- Epistemological frames
- Abductive reasoning
- Change in complex systems
- Interdependency
- Improvisation/sense-making.
All told, the idea of ‘warm data’ reminds us of the value of using what Bateson terms ‘other species of information’ in education, in order to foreground and better understand new patterns of connection not visible though current data practices and routines.