Studying attention to IPCC climate change maps with mobile eye-tracking

surface temperature, sea level rise, anthropogenic CO2 emissions, and atmospheric particulate matter concentrations (PM2.5). These maps were generated using data visualisation tools [27, 28], provided by the Intergovernmental Panel on Climate Change (IPCC) and were accessible at interactive-atlas.ipcc.ch at the time of writing (see S1 Fig for an overview). The supplemen- tary stimuli for Part 2 included two large-scale artworks printed on A0-sized posters and mounted on the laboratory wall (see S2 Fig for details). The Climate Change Anxiety Scale (CCAS), a 22-item questionnaire [29], was used to mea- sure participants’ responses to climate change on a 5-point Likert scale, covering four subcate- gories (see S1 File for the full scale). An exit-questionnaire was also administered to gather optional demographics data and participant feedback. Gaze data were recorded with the Pupil Invisible mobile eye-tracker using Pupil Invisible Companion App (version 1.4.21). The raw gaze data were pre-processed on the GDPR-com- pliant Pupil Cloud platform. The maps were presented using PsychoPy (version 2022.2.5), the CCAS and exit-questionnaire were presented using Google Forms (see S3 Fig for the experi- mental setup). 2.3 Design This study was primarily exploratory and descriptive, aiming to analyse participants’ viewing patterns using mobile eye-tracking data while they engaged with ten world maps depicting the climate crisis and two artworks in a laboratory setting. Data were collected under two viewing conditions: a primary individual viewing condition,

where participants viewed the stimuli alone, and a secondary paired viewing condition, where pairs of participants viewed the stimuli together. Despite challenges in recruiting pairs, which resulted in unequal group sizes (35 and 12 in single- and paired-viewing conditions) the setup allowed for the collection and comparative analysis of gaze metrics across different viewing contexts. The primary analyses focused on fixation-based metrics and their derivatives, which are often linked to attentional processes, such as overt attention and visual attention guidance. Descriptive statistics such as averages, frequencies, and heatmap visualisations were used to present data rather than formal hypothesis testing, highlighting engagement patterns with the stimuli. Additionally, the study served as a preliminary evaluation of the feasibility of the research procedures and the analysis pipeline in a laboratory setting and beyond, establishing ground- work for future in-situ experiments in public spaces. A significant goal was to compile a size- able eye-tracking dataset, which, due to the inherent limitations of mobile eye-tracking systems, was expected to be noisier than data from stationary systems. Lastly, this open dataset and code were prepared for reuse in future research, allowing for expanded analyses. 2.4 Procedure The experiment was conducted during regular working hours. Participants began by receiving detailed written and oral instructions, and the experimenter addressed any questions. They

provided written informed consent before being equipped with the mobile eye tracker, with correction lens inserts provided when needed. Although the mobile eye tracker was calibra- tion-free and self-correcting, calibration was visually checked using a standard five-point cali- bration panel, and offset corrections were applied if necessary. For the first part of the experiment (map viewing), participants’ calibration was confirmed before the recording started. Each participant was assigned a randomly generated three-digit ID and viewed maps displayed on a laptop in a semi-randomised order, spending at least 30 PLOS ONE Studying attention to IPCC climate change maps with mobile eye-tracking PLOS ONE | https://doi.org/10.1371/journal.pone.0316909 January 10, 2025 4 / 17