System Dynamics (SD) modelling supports decision-making by simulating and projecting key
performance indicators (KPIs) of a system. To assess uncertainty in these KPIs, modellers
typically vary model parameters and run multiple simulations. This process illustrates potential
scenarios and generates KPI distributions that provide quantified measures of uncertainty
through statistical analysis.
However, parameters often do not vary independently. Studies across sectors have reported
correlations and dependencies among parameters (Krefeld-Schwalb et al., 2022; Li and Vu,
2013). When varying parameters in SD models, whether these dependencies are accounted for
can shape the combinations of parameter values, influence the distributions of projected KPIs
and the derived insights. This issue has not been thoroughly addressed in the SD literature.
To highlight the importance of parameter dependence, we present a copula-based experiment,
modelling dependencies between SD model parameters in several different ways and
comparing the resulting KPI distributions. The experiment demonstrates that both the strength
of correlations and the structure of dependencies can affect KPI uncertainty. These findings
motivate the adoption of more flexible approaches to adequately model such dependencies.
The main contribution of this thesis is a method that models dependencies among SD model
parameters using Bayesian Networks (BNs) to improve KPI uncertainty assessment. BNs
provide a flexible framework and algorithms for uncovering complex conditional relationships
from data and integrating them with expert knowledge. We apply the approach to an epidemic
SD model, where dependencies among epidemiological parameters are estimated from a crosscountry COVID-19 dataset using a BN and validated against domain knowledge. The learned
BN produces input for analysing KPIs of the epidemic SD model and yields uncertainty
envelopes that differ from those generated by independent or single-copula priors.
This study offers SD practitioners a practical way to incorporate empirically grounded multiparameter dependencies into their models, enhancing the defensibility of uncertainty
assessments while keeping additional data-collection effort manageable. The proposed method
contributes to both the SD and mixed-methods research literatures.
| Date of Award | 6 Feb 2026 |
|---|
| Original language | English |
|---|
| Awarding Institution | - University Of Strathclyde
|
|---|
| Sponsors | University of Strathclyde |
|---|
| Supervisor | Susan Howick (Supervisor) & John Quigley (Supervisor) |
|---|