Royal Geographical Society-Institute of British Geographers conference, 2013

Three papers based on the project were presented at this year’s RGS-IBG conference. These are:

In the ‘Researching Practices for Sustainability’ stream
Using Qualitative Data to Model Practices: new possibilities, limits and challenges (presented by James)
This paper reflects upon current work funded under the ESRC Secondary Data Analysis programme. The project, Changing Commutes, runs between February 2013 and July 2014 and is developing an agent-based model exploring how social influence and social learning shape the uptake of cycle commuting, conceptualised as a practice rooted in (and itself shaping) cultural, infrastructural, and organisational contexts. The project re-uses qualitative data to develop rules for the behaviour of individuals (and organisational actors) within the model. These agents are seen as learning through interactions with others, within socio-spatial networks.
The talk discusses the use of the qualitative data within the project, exploring the challenges involved in re-coding data from a range of sources and using it to develop behavioural rules. The qualitative data used is mainly interview transcripts, from projects related to cycling, including interviews both with cyclists (and non-cyclists) and with stakeholders/policy-makers. Most are individual interviews but some are focus groups/group interviews, and the data corpus also includes some workshop and ethnographic data. I led one of the projects from which data is used, but was not involved in other projects. There are many challenges (and limitations) involved in using qualitative data in this way; however, the paper argues that this innovative use of data can provide new insights into how and where tipping points in the uptake of more sustainable practices might occur.

In the ‘Mobility as Practice’ stream
Practice into Practice: (how) can we use practice theory to inform modelling ‘behaviour change’? (presented by Rachel)
This paper will ask whether and how practice theory might be operationalised for modelling. It reflects upon current work funded under the ESRC Secondary Data Analysis programme. The project, Changing Commutes, runs between February 2013 and July 2014 and is developing an agent-based model exploring how social influence and social learning shape the uptake of cycle commuting. It is re-using qualitative data to develop rules for the behaviour of agents within the model, conceptualised initially as individual actors. Many agent-based models have conceptualised people as rational actors who make decisions on the basis of utility; however, this does not have to be the case and this is not how we intend our model to work.
But is an approach that models individuals by its nature antithetical to practice theory? Or can it be developed in a way that fits within a practice theory approach? The latter might involve, for example, creating an internal structure to the agents such that they accumulate ‘competences’, while policy interventions might shift the extent to which such competences are located in agents themselves. This paper will also consider other possible approaches, including using agent-based modelling to model organisations, or using more holistic modelling methods such as system dynamics modelling.

In the ‘Modelling for Policy’ stream
Using Qualitative Data to Develop Rules for Agent Behaviour: lessons from the Changing Commutes project (presented by Rachel)
Researchers are seeking to develop more sophisticated methods of modelling agent behaviour, and to use new forms of data in building such models. This presentation will reflect upon the Changing Commutes project, which is funded under the ESRC Secondary Data Analysis programme. The project, which runs between February 2013 and July 2014, is developing an agent-based model exploring how social influence and social learning shape the uptake of cycle commuting.
The talk discusses the use of the qualitative data within the project, exploring the challenges involved in re-coding data from a range of sources and using it to develop behavioural rules. The qualitative data used is mainly interview transcripts, from projects related to cycling, including interviews with cyclists (and non-cyclists) and with stakeholders/policy-makers. Most are individual interviews but some are focus groups/group interviews, and the data corpus also includes some workshop and ethnographic data. I led one of the projects from which data is used, but was not involved in other projects.
Issues include the combination of different types of data and the comparability of the contexts, as well as the extent to which a researcher not involved in the original project is able to meaningfully recode the data. The presentation will discuss ways in which we can develop the use of different forms of qualitative data in modelling, as well as the limitations of qualitative data for this purpose and the most suitable forms of data for different modelling purposes.

Comments are closed.