Page 145 - Demo

Discussion and Conclusions1436also lead to insights into the topics visualisations circulating within their context. Thus, Chapter 5 states that the circulation of visualisation can network topics and that observing the nature of this connection can be relevant when investigating the networked topics. The chapter suggests, for example, that the nature of the visual networking of nanotechnology in food and GMOs is worthy of investigation, as it may lead to more knowledge about these two topics.Third, a change in the meaning of visualisations can take place when a topic connects visualisations together, and over time some visualisations are disconnected from the network while new visualisations are connected to it. The studied online fracking controversies seem to suggest that the meaning of networked visualisations changes over time because different types, content and narratives of visualisations are networked. This type of change in the meaning of visualisations was previously studied in the context of climate change (Niederer, 2018; Niederer & Pearce, 2017), comparing the platforms on which visualisations are networked. This thesis suggests studying this type of change to gain knowledge of the role of visualisations in the dynamics of online policy controversies.6.3 Reflection on, and discussion of, methods and recommendations for further researchIn this dissertation, I took an interpretive approach and further developed it to also include the visual. I chose an interpretive approach as this dissertation focusses on visualisations as objects making meaning, and the interpretive approach aims to understand how different meanings given to human-created artefacts by various (groups of ) people exist (Schwartz-Shea & Yanow, 2012, p. 42; see also Yanow, 2007). In this section, I reflect on the methods used for gathering and analysing data and I point out a number of directions for further research.6.3.1 Limitations to working with digital methodsThroughout the project, I worked with several digital methods tools and each time found that this comes with different limitations. For example, there are limitations to the reliance on commercial companies (Google and Twitter). There are also limitations to automated categorisation software of both visuals (ClarifAI) and text (sentiment analysis as developed by Roel Kuiper, a student in the team Efrat.indd 143 19-09-2023 09:47