What’s the impact of impact?

I stumbled into impact. In my research on crowd behaviour, I was interested in theory primarily, though I could see and sought to combat those theories (such as that of Gustave Le Bon's) that weren't just factually wrong but dangerous in practice. But my own 'practice' in psychology was limited to writing and speaking at conferences. With Stephen Reicher and Chris Cocking, I started research on behaviour in mass emergencies, and I found that practitioners and policymakers were interested in the practical implications of the work (see Drury, 2016). I also found that I enjoyed sharing the research with these non-academics – crowd safety managers, fire service personnel, and civil contingencies planners – and discussing with them how to solve their practical problems.

That was a bit before the 'impact agenda' was a thing in UK psychology. Later, my experience of impact work was useful when I took on roles as Impact lead and then Director of Research and Knowledge Exchange in my School at the University of Sussex. I have been working with colleagues to help them understand and achieve impact, and I have been lucky enough to witness some impact success stories.

All of this has enabled me to see not only the way that psychology research can contribute to changes in the world outside of academic – which is the standard definition of 'impact' – but also the way that achieving impact can itself have consequences for the practice of psychology. Based on my experiences, here I discuss three areas where impact can impact on psychology.

Validation and theory development

Kurt Lewin's (1945) famous phrase 'there is nothing as practical as a good theory' (p.129) argues for the value of theory that can solve practical problems. But applying theory to practical problems can also provide an opportunity for the validation and development of theory. 

If the practitioners are applying the theory as specified, and the outcome is as predicted, then this provides new evidence for the theory, using a method or dataset different than that of the experiments, observations or interviews that were the original evidence base for theory. An example I use in my teaching is Clifford Stott's 'natural experiment' during the Euro2004 football championships (Stott et al., 2008). Based on ethnographic and observational studies and on the elaborated social identity model, Clifford and Otto Adang had developed a hypothesis that 'disorder' in football crowds emerges when fans perceive policing as disproportionate and when that policing relies on coercion.

Fans experience this policing as illegitimate, which makes conflict more acceptable and possible to them, and indiscriminate, which makes conflictual fans ('hooligans') become seen as ingroup. But if policing is experienced as 'proportionate' and facilitative of the fans' aims (of enjoying the game), there will be lower levels of conflict and fans will collectively self-regulate based on norms of enjoying football. 

Portugal has two police forces: the GNR (responsible for policing rural areas and small towns) and the PSP (responsible for policing cities). Only the PSP accepted the model and wanted to use it to inform their practices. For the football tournament, Stott and Adang assembled a large research team to gather a wide variety of types of evidence: interviews, observations, surveys and ethnographic. 

The different sources converged in support of the hypothesis. For example, in the GNR areas there were two major riots and 52 arrests, while in PSP areas only 0.2 per cent of observations were of conflict and there were no police records of major collective disorder. This study of the application of theory is perhaps the most powerful body of evidence in support of the hypothesis on the role of policing in football crowd conflict. 

The above example relates to validation, but working with practitioners can also be a way to develop a theory. When people ask me for key advice on impact, my short answer is to develop relationships with relevant non-academics. The development of relationships with 'end-users' or 'stakeholders' (i.e. those practitioners or policymakers outside academia who might find the research/theory useful) is the essence of what is sometimes called the pathway to impact. Once a relationship has been developed, it can enable a dialogue over the implementation of the theory which can in turn be an opportunity to refine or extend the theory, or specify some of its boundary conditions. 

An example is the training that Holly Carter and I provide for the UK fire and rescue service. We provide regular sessions as part of the training for firefighters involved in mass casualty decontamination in chemical, biological, radiological and nuclear incidents. Until recently, training, exercises and guidance for mass decontamination either ignored psychology (and focused on technical aspects of the decontamination process) or drew upon a discredited psychology of automatic public panic and disorder. 

Holly's research (Carter et al., 2015) showed that any such public anxiety and disengagement was not natural and inevitable, but rather was a function of the way the decontamination process was managed by responders. The alternative to shouting, attempts at coercion and lack of information (which were the factors that people said distressed them during decontamination and led to disengagement with the process) was simple: communication. In short, explain the rationale for the process; explain how it works; explain how to remove contaminants from your body; show understanding for people's concerns, even if you can't meet all of them. 

Presenting the research and the recommendations as part of the training sessions has been an opportunity for questions from practitioners, and for thinking about the gaps that could be addressed by further research. For example, some staff said they would find a script useful, so a strand of research began examining the effectiveness of different forms of words. There were also questions about how communication could work when some people would not listen to professional responders; a solution would be to get other members of the public to engage with them.

This is a hypothesis derived from other contexts (crowd safety management at live events) and could be tested systematically in the mass decontamination context. One issue to consider is whether the practitioners are really implementing your model accurately, or some simplified (even distorted) version of it. The solution here is to work closely with them (in order to observe and help in evaluation), not just leave them to get on with it.

Student interest and demand

A second way that impactful research impacts back upon the discipline of psychology is through teaching. In social psychology, in particular, students like research and theory with real-world applications. This means for those researchers involved in impact activity, examples from their own work can be included in lectures and in seminar activity. One technique I use is to structure part of a seminar around a real-world problem related to my research or impact activity and ask the students to try to solve it.

For example, in one seminar the task for students is to improve egress times when there is a fire alarm. If the real behavioural risk in fires is not panic but people reacting too slowly to fire alarms, how can this be addressed? What measures would the students advise in drills and preparation? In short, impact provides an opportunity for adding value to undergraduate teaching, by providing interesting content and engaging activities. Ideally, the students not only engage with and understand the material better, but also appreciate how they can apply psychological concepts themselves in the world outside the classroom.

Being involved in impact activity can also serve to attract PhD students in the same way that a research profile does. Many students are attracted to advanced study and research in psychology are doing research because they want to make a difference. They want to work with researchers where they see real-world problems are being addressed and they develop research proposals that likewise address these problems.

Process thinking

The third way that impact outside of academic psychology can impact psychology itself is through encouraging process thinking. The context here is the impact agenda which has changed the way we think about 'applied' work. In my view, there are a number of limitations of the impact case study format that was part of REF2021. 

It led too many researchers to think that if they didn't have enough evidence or impact to produce a 3* or 4* impact case study, then their impact activity wasn't worth doing or documenting; and it placed too much weight on a small number of impact examples and often a relatively small number of individuals (in fact, most of the unambiguously 4* impact case studies in REF2021 were team efforts). But the impact case study format, and the sections on impact (now modified) for grant applications to UKRI councils, have helped people to understand impact as a process. 

Who outside of academia would find this research interesting, useful, or necessary? Could you involve those stakeholders in the development of the proposal, rather than simply approach them when you have the results, to ensure that there is a connection and continuity between research and impact? How will you engage with people outside of academia – meetings, reports, videos? What are the strengths and weaknesses of each as a way of disseminating and discussing the research? What kind of language is needed to communicate the research to these non-academic audiences? Finally, how will you document and record any impacts your research achieves, particularly considering that impact might not occur until years after the original research? 

Before the 'impact agenda' was a thing, impact was a happy accident for many of us. Now we tend to think more strategically about it. This is not (or should not be) driven just by a bureaucratic need to measure and assess. If we think our work can have positive benefits outside of academia, then this process thinking can help us maximise these benefits, as well as provide opportunities to test and develop our models and engage our students.

John Drury is Professor of Social Psychology at the University of Sussex.
[email protected]

Key sources

Carter, H., Drury, J., Rubin, G.J., Williams, R. & Amlôt, R. (2015). Applying crowd psychology to develop recommendations for the management of mass decontamination. Health Security, 13(1), 45-53. doi: 10.1089/hs.2014.0061
Drury, J. (2016). Impact: From riots to crowd safety. The Psychologist, January. https://www.bps.org.uk/psychologist/impact-riots-crowd-safety 
Lewin, K. (1945). The research center for group dynamics at Massachusetts Institute of Technology. Sociometry, 8(2), 126-136.
Stott, C., Adang, O., Livingstone, A. & Schreiber, M. (2008). Tackling football hooliganism: A quantitative study of public order, policing and crowd psychology. Psychology, Public Policy, and Law, 14(2), 115-141.