Table 1. Evidence-based medicine versus complex systems research paradigms.

Adapted under Creative Commons licence from Greenhalgh and Papoutsi [16].

	Evidence-based medicine paradigm	Complex systems paradigm
Perspective on scientific truth	Singular, independent of the observer, ascertainable through empirical inquiry	Multiple, influenced by mode of inquiry and perspective taken
Goal of research	Establishing the truth; finding more or less universal and generalisable solutions to well-defined problems	Exploring tensions; generating insights and wisdom; exposing multiple perspectives; viewing complex systems as moving targets
Assumed model of causality	Linear, cause-and-effect causality (perhaps incorporating mediators and moderators)	Emergent causality: multiple interacting influences account for a particular outcome but none can be said to have a fixed ‘effect size’
Typical format of research question	“What is the effect size of the intervention on the predefined outcome, and is it statistically significant?”	“What combination of influences has generated this phenomenon? What does the intervention of interest contribute? What happens to the system and its actors if we intervene in a particular way? What are the unintended consequences elsewhere in the system?”
Mode of representation	Attempt to represent science in one authoritative voice	Attempt to illustrate the plurality of voices inherent in the research and phenomena under study
Good research is characterised by	Methodological ‘rigour’, i.e. strict application of structured and standardised design, conventional approaches to generalisability and validity	Strong theory, flexible methods, pragmatic adaptation to emerging circumstances, contribution to generative learning and theoretical transferability
Purpose of theorising	Disjunctive: simplification and abstraction; breaking problems down into analysable parts	Conjunctive: drawing parts of the problem together to produce a rich, nuanced picture of what is going on and why
Approach to data	Research should continue until data collection is complete	Data will never be complete or perfect; decisions often need to be made despite incomplete or contested data
Analytic focus	Dualisms: A versus B; influence of X on Y	Dualities: inter-relationships and dynamic tensions between A, B, C and other emergent aspects