5. Study Design – Focus Groups
In light of the questions we pursue about designing for public sonification, we constructed a focus group study and conducted two sessions, with a total of 12 participants, focusing on just two recordings of the four Canadian cities: Toronto and Vancouver. Rather than individual listening tests aimed primarily at determining the perceivability of sonified information, we wanted the flexibility of some individual tasks but also the open-endedness of focus group brainstorming in order to elicit a more natural group reaction to hearing socially relevant yet specialized air pollution data. For the group discussion we wanted to engage participants in thinking about the role that popular infographics and visualizations play in their informational ecology, ultimately brainstorming about how sonifications might fit in. Rather than using a random sample we were interested in more informed opinions, so for this directed group discussion we selected practitioners who work or study in related communication and visualization fields. The participant group involved five men and seven women, ranging in age from early 20s to late 40s, all adults with normal/average hearing and no special musical training. The focus group followed a modular structure of group and individual activities. First, the groups listened to two examples of air pollution in Canadian cities (one month – January 2014 – compressed into one minute of audio for Toronto and Vancouver, respectively). In this task participants were given no explanation and asked to comment on which city was more polluted and why they thought that, including a more general discussion on what “pollution” ought to sound like. Next, groups were offered a brief explanation of the air pollutants sonified and the mappings used, and each participant was given a worksheet (Fig 2) asking them to collectively listen and individually complete three short tasks.
The first task involved following along on a visual graph while listening to the values of ozone over 12 days, compressed into 1 minute of audio. The second task involved listening to a combination of carbon monoxide and sulphur dioxide, sonified together at the same resolution – 12 days into 1 minute of audio – in order to identify which of the two is portrayed on a visual graph. The third task involved listening to a mix of carbon monoxide and ozone sonified together; while the ozone graph was provided, participants had to graph carbon monoxide levels by hand. The motivation for this activity was to provide opportunity for individual listening evaluation within the format of a focus group and to engage participants in the constructive task of graphing as well as listening in order to generate a more rich discussion of the way participants perceived detail in the sonifications. Given the aim of the study – public perception rather than scientific analysis – our goal here was not to confirm hearability in the model but to probe how everyday listeners listen to data, which barriers to deeper understanding present themselves, and whether sonified data generates further interest in the social issues they may represent. Of course, we are aware that testing formats influence possible biases and that by using visual graphs we create certain limitations; however, the proliferation of common visualizations in the public domain offers an opportunity to probe how these two modalities complement (or not) each other in comprehension and reception of data by the public.
