Understanding and shaping complex social psychological systems: Lessons from an emerging paradigm to thrive in an uncertain world

Abstract

In today's rapidly evolving world, human behaviour plays a crucial role in addressing complex challenges. These span from mitigating non-communicable diseases, to climate action and pandemic preparedness. Traditional behaviour change research has focused on identifying and addressing specific factors that impede positive change, using a decomposition-based approach. This approach breaks key behaviours down into their component parts, and seeks to affect influences – again broken down into e.g. attitudes, social norms, resources and other opportunity-related factors – on these components. Component-dominant dynamics refer to a situation, where the behaviour of a system is determined primarily by the properties of its individual components, rather than by emergent phenomena. The decomposition-based approach is highly effective when this is the case. However, the approach may have severely limited effectiveness in contexts characterized by interaction-dominant dynamics, where outcomes are determined not by individual components, but emerge from the ongoing interdependent influences between them. Interaction-dominant systems have been studied in various fields under the interdisciplinary rubric of complex systems science. This work argues that people are active agentic creatures, who are self-determined and self-organising experts of their own environments, which infuses social systems with inherent non-stationarity and hence uncertainty: Objects of study in behaviour change science, as well as their relationships, change. In addition, small events can cause large impacts, and long periods of apparent stability can be punctuated by rapid change. This implies that past data may be of limited use for inference and interventions across contexts or time. Indeed, the opportunities and risks laying in the future of complex systems or decision making contexts, are vastly more numerous than those which could be called merely complicated. Therefore, it should be carefully evaluated, to what extent a given situation is indeed amenable to decomposition-based solutions. While it might be impossible to predict the long-term future of some complex “non-linear” systems, risks and opportunities can be evaluated based on evolutionary potential of their space of possibilities. The specific contribution of articles included in this work supports the argument as follows. Article I takes the view of people as agentic, autonomous decision makers, and points to the need to foster their capacity to self-organise. It describes a compendium of techniques, which (combined with proper scaffolding and support) could enable individuals to better self-manage their motivation and behaviour. Article II uses data containing self-management techniques to demonstrate a conceptually important model, where behaviours – and what the literature has conventionally considered their influences, precursors, or determinants – are represented as components of a mutually interacting network. The network representation reflects a deviation from the conventional conceptualisation, which hinges on component-dominant dynamics to depict simplistic causes and their effects with boxes and arrows. The article also discusses the importance of distributional shapes, as well as how summary statistics developed for simple, symmetric distributions can misdirect inference when the shape is more varied. Article III proposes a process definition of behaviour change, particularly calling for attention to some core features of complex systems; interconnectedness, non-ergodicity and non-linearity. It points out how interaction-dominant dynamics can produce distributional shapes, poorly amenable conventional analysis. It also discusses problems of traditional linear models of between-individual data for studying behaviour change, and expands the idea of aforementioned between-individual networks to those containing temporal recurrences of idiographic system states – attractors. Article IV depicts how the conceptualisation of behaviour change as movement in an attractor landscape can be used to understand change on different scales of observation, from individuals to communities and the society at large. It brings our attention to how non-linearities in interconnected systems can mean sudden transitions after long periods of stability and, more generally, the need to be wary of attractor states yet unseen, when there is uncertainty about distribution shapes.To address the transparency and irreducibility concerns currently being voiced across scientific disciplines, all the data and analysis scripts are made publicly available online. Apart from Article I, these supplements are produced directly from the data as accompanying websites to reduce error and increase accessibility.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.