In your research on urban data and visualisation, you have coined the term 'seamful visualisation' as a design technique to show uncertainties and gaps in data. How can designers and data journalists use data visualisations to foster civic purposes and promote public literacy of data?  

Data journalism and civic activities involving (urban) data, like citizen science, have been emerging as empowering tools, practices and methods to negotiate or raise awareness about issues relating to cities. Academic discourses discuss a lot how data are always partial, how they are situated, local and contextual representations of phenomena. When you become familiar with specific data sets and how they represent a specific phenomenon, the limitations of these data become more obvious. However, where visualisations aim to communicate pertinent concerns to broader publics, including the interpretations of citizens and non-experts, I think that the graphic representation of these data should be carefully considered. That is where seamful visualisation techniques come into play.  

Seamful visualisation techniques have been tentatively mentioned in previous research and have been developed as a strategy for visualisation design that complicates how data, and thus phenomena, are represented. Outside of visualisation research, seamfulness has been explored in more depth in HCI and ubiquitous computing research as a strategy to productively engage with the ‘seams’ between different systems, or between a virtual world and reality.   

In my PhD research, I extended this approach and explored seamful visualisation as a critical practice. I explored visualisations of geospatial data to reveal more ambiguities about how data represent and relate to phenomenon in physical space. This seamful approach to visualisation design integrates concerns about the broader limitations and qualities of data that are often raised in critical data studies and critical visualisation research. Data are always situated and partial, in a way, and this needs to be reflected in a visualisation.  

The seamful approach to data visualisation focuses on depicting these broader data limitations and qualities within visualisations by relating data to the phenomenon. In effect, this can chart the relationship between data and what they represent, and by doing so, we can highlight, for example, missing data, uncertainties, partialities, biases or definitions inherent to data. These inconsistencies only emerge when considering the real-world phenomenon side by side with the data used to represent it. My approach investigates how data represent what they represent—what aspects of a phenomenon are and are not included in a (visual) data representation—instead of taking a data set at face value.   

In my research, I explored this through design explorations that don’t necessarily produce visualisations that can be dispersed as stand-alone and self-explanatory visualisation objects (e.g. graphics). I believe there is value in exploring seamfulness in more conventional visualisation practices as they are used in journalism, for example. When visualisations aim to address publics and civic purposes, visualisation authors can consider actively including information and visual cues about the qualities and limitations of data. This is already sometimes done when it comes to limitations that are an explicit part of the data, like missing data or uncertainties of specific values. However, to communicate data in a way that stresses their foundational limitations, partialities and situatedness, broader and more implicit limitations should be considered for visualisation designs. This seamful approach can highlight what parts of a phenomenon are not represented in the data in the first place, e.g. due to implicit definitions inherent in the data. For visualisation authors to be able to consider and depict such limitations they need to closely engage with the data and their limitations to build rich knowledge about a ‘given’ data set. The resulting visualisations can enable critical engagements with data by providing richer representations of data and phenomena to viewers.  

How do infrastructures matter for citizens?  

In the context of my research in critical data visualisation and critical data studies, I mainly consider infrastructures related to data: the socio-technical elements that enable processes of data generation, storage, analysis, dispersion, and communication. From this perspective, a range of (public) infrastructures are required and relied on so that citizens are able to engage with data and visualisations in journalistic or public environments. This includes infrastructures enabling data collection, like sensor networks, and includes software and platforms that enable analysing, visualising and communicating data.  

How should knowledge about infrastructures be communicated and what role can journalists, researchers or public officials play?   

Data are increasingly used to communicate and present issues to citizens, for example in the contexts of journalism or open data initiatives. But how data infrastructures are involved in these practices and how they enable, limit and influence what is collected (or not collected) as data and visualisations is usually not proactively exposed. I think that fostering critical public understandings and literacies of data and their visualisation, the contributions and limitations of specific (data) infrastructures should be openly communicated. Infrastructures involved in data and visualisation practices can be interrogated and opened up for public consideration. It is important to consider how a phenomenon is represented in data and how a visualisation is used. Every visualisation of data involves a range of steps through abstraction, simplification, reduction and amplification. These journalistic or design processes need to be made more transparent in some way to enable a better (public) understanding of how data come to be, how they represent phenomena, and what insights that can be taken from them.