Bayesian network analysis of domestic water and energy use in Bristol
La Matta Romero, Fiorella; Staddon, Chad; Adkins, Deborah; Lewis, Todd
Authors
Fiorella La Matta Romero
Chad Staddon Chad.Staddon@uwe.ac.uk
Professor/Associate Head of Department: Research and Scholarship
Dr Deborah Adkins Deborah.Adkins@uwe.ac.uk
Wallscourt Fellow in Sustainable Buildings
Todd Lewis Todd.Lewis@uwe.ac.uk
Senior Lecturer in Ecology and Environmental Technology
Abstract
Understanding the factors influencing domestic water and energy conservation is crucial for promoting sustainable development and resource conservation while achieving well-being and comfort at home. Bayesian networks (BN) are a powerful tool for modelling probabilistic dependencies among variables, allowing researchers to uncover hidden patterns and causal relationships. With their ability to express parameter uncertainty through probability distributions and reflect the dependencies of each variable, Bayesian networks are a valuable resource for researchers seeking to understand complex systems. This study introduces a novel Bayesian network analysis to model the intricate correlations between household demographics, property characteristics, and psychological factors in determining water and energy usage patterns, based on a case study in Bristol developed through an innovative approach called “HomeLabs”. Utilising survey data from 24 households, variables such as age, marital status, gender, income, education level, dwelling type, and the presence of utility meters were assessed alongside psychological attributes such as awareness of resource use, conservation motivations, and behavioural patterns. By modelling the probabilistic relationships among these variables, the Bayesian network provides insights into the complex dynamics of household consumption. Preliminary insights indicate that certain demographic and property variables, when mixed with behavioural factors, are significant drivers for the efficient use of resources. This study provides a comprehensive framework for policymakers and stakeholders to develop tailored, effective conservation intervention strategies. Identification of the primary factors influencing resource use behaviour will pave the way for targeted measures that promote sustainable practices tailored to the specific needs of Bristol communities. This study not only extends the frontier of research methodology in socio-technical systems but also provides actionable insights for promoting conservation behaviours at the household level, crucial for mitigating environmental impacts in urban communities like Bristol while promoting more mindful consumers.
Presentation Conference Type
Poster
Conference Name
Royal Statistical Society 2023 International Conference
