Skip to content

Blog post

Complex realism and educational research

Selma Dogan, Doctoral Researcher at Cardiff University

My current ESRC-funded PhD research crosses the boundaries of linear thinking by drawing on ‘complex realism’ – the conceptual framework through which I investigate the mechanisms that lead mature undergraduates to withdraw from their degree studies in Welsh universities.

There is more to reality than meets the eye. As Malcolm Williams explains, reality has a contingent nature (Williams, 2021). The probability of an event A is contingent and conditional upon an event B (Williams, 2011, p. 41). Using this contingency and the ontologically known stable entities in our social lives where we perform our duties and responsibilities while observing certain social rules, we are able to predict the unknown (Williams, 2015, p. 89). In the natural world, there is natural necessity – that is, some things must happen (the sun rising in the east), and some things are impossible (no human can fly unaided to the moon).

But the social world is steered by the probability of occurrence (Williams, 2021, p. 40). All events in the social world have a probability of greater than zero and less than one – for instance, you will take the train to work (if it is the workday and you usually prefer the train) in which case, the probability is very close to 1; or you will be the next king or queen of Britain, the probability of which is very close to 0 unless you are somehow related to the Royal family.

Central to ‘complexity’ is the idea of ‘non-linearity’. Educational systems are represented by multiple variables connecting in non-linear ways in which few variables may interact with many, and many with few. Warren et al. (1998) discuss the non-linear ways of learning in detail: in the beginning, ‘learning the ropes’ is slow. Initially, we do not even realise that we are learning, and we feel discontent with our progress. As time goes on, we accumulate knowledge making connections between our prior educational experiences and new learning. Our learning gains impetus until a moment comes, and we bump up against resource limitations, perhaps, due to a pandemic, a financial hardship, or a lack of adequate learning resources.

Complexity theory is framed by two main descriptors contextualising the real: ‘phase space’ and ‘strange attractors’. Phase space is part of the wider context of social reality with overlapping layers of history, current time and future and where anything can happen but not everything will, given the rule-driven nature of the social world (Williams, 2021, p. 43). The layers of time are inherently interdependent, generating a complex network of interactions in educational systems – the relationship between the student and the lecturer is defined by many personal and institutional dynamics which do not come into existence in a vacuum.

‘The relationship between the student and the lecturer is defined by many personal and institutional dynamics which do not come into existence in a vacuum.’

In my research, the mechanisms which trigger non-continuation in mature students’ trajectories, are a set of time-ordered and ‘strange’ attractors from a wide range of ‘initial conditions’. These ‘strange’ attractors emerge as a result of a slight change in the context. They are the patterns shared by a number of cases within the phase space and towards which other student trajectories converge during complex interactions (Williams 2021, p. 43). A ‘strange’ attractor may be the lack of effective communication between the enrolment process and a mature student in a Welsh university.

Complex realism offers a generative model of causality (see figure 1) as it exposes the initial conditions in students’ trajectories (Williams, 2021).


Figure 1: A generative model of mechanisms

In figure 1, the underlying mechanisms, which the interviews with the students in my research will identify, connects the two events (X and Y) and the context in which the relationship occurs. The probabilities of those initial conditions which can take me to the underlying mechanisms will be estimated via the surveys I will initiate, to infer a causal relationship between X and Y.

During this research, complex realism allows me to break the habit of forming linear perceptions of the social and academic contexts, and to generate generalisable and valid findings around the mechanisms for mature student withdrawal.


Warren, K., Franklin, C., & Streeter, C. L. (1998). New directions in systems theory: Chaos and complexity. Social Work, 43(4), 357–372.

Williams, M. (2011). Contingent realism: Abandoning necessity. Social Epistemology, 25(1), 37–56.

Williams, M. (2015). Situated objectivity, values and realism. European Journal of Social Theory, 18(1), 76–92.

Williams, M. (2021). Realism and complexity in social science. Routledge.

More content by Selma Dogan