Visual art inspired by climate change—An analysis of audience reactions to 37 artworks presented during 21st UN climate summit in Paris

above). The participants were asked if they imagined the artist to be (1) someone like them- selves, (2) someone with values similar to themselves, and (3) someone expressing the values of the public. As indicated above, these items were apparently measuring relatively different components, so conclusions with respect to the role of identification with the artist should be drawn with caution. The items were answered on a seven-point scale (1 = strongly disagree, 7 = strongly agree). Reflections on the artwork. This variable consists of six components capturing the degree to which the artwork made the audience reflect on climate change: (1) The artwork makes me think and reflect on its meaning, (2) the artwork seems relevant to my daily life, (3) the artwork highlights consequences of climate change that would affect me personally, (4) the artwork makes me think about the problem of climate change, (5) the artwork makes me think about my own role within the current climate situation, and (6) the artwork makes me more aware of my behavior’s impact on the environment. The items were answered on a seven-point scale (1 = strongly disagree, 7 = strongly agree). Support for climate policy. As dependent variable, we decided to use the priority of cli- mate policies expressed in a longer block on policy priorities as an answer to the item: How important is it to you that climate change and the environment are given a high priority in pol- icymaking? The answers were (1) it is rather unimportant in comparison, (2) it is not that

important to me, (3) it is important but not the highest on the list, and (4) it should have the highest priority. This item was presented in a block of different policy priorities (e.g., economic prosperity, security, etc.) to give it a context. The other policy priorities are not analyzed fur- ther in this paper. Control variables. To control for confounding influences of different compositions of audiences, the following variables were included in all analyses: Age, gender (0 = female, 1 = male), environmental attitude (measured by the item: "Comparing yourself to others, how interested would you say that you are in environmental issues?" 1 = far below average; 4 = about average; 7 = far above average), and experience with art (as introduced above in the description of the sample). In the initial analyses, age and experience with art did not relate to the model variables, so they were excluded from further analyses. Artwork characteristics. For each artwork, one to two members of the research team recorded characteristics describing the artwork, for example, the type of artwork (painting, sculpture, photograph, etc.), the materials used, the size, if it was an artwork with participatory elements if it depicted familiar objects, animals, humans, etc. Also, basic features of art percep- tion psychology were captured (e.g., if the artwork made use of the peak shift principle). These characteristics were used in the artwork level analyses described below. Analysis strategy The data has a nested two-level structure since the audience responses have been recorded at

37 venues for artworks in Paris, with a number of people per artwork observing the same art- work. This allows for simultaneous tests of effects of the artwork (e.g., how features of the art- work impact the emotional response it evokes) and effects of person-level factors (e.g., the link between the strength of an individual emotional response and perceived quality). SEM soft- ware MPLUS 8.3 was used for the analyses [54]. The COMPLEX analysis type in MPLUS was used, adjusting standard errors for effects of clustered / nested data. Since the research question of this study taps into new ground, we follow a partly confirma- tory (step 1), partly exploratory strategy in the data analysis (step 2 and 3): 1. After the initial confirmatory factor analysis (see supporting information), we ran a confir- matory model test on the structural model as has been derived theoretically in the PLOS ONE Visual art inspired by climate change — An analysis of audience reactions PLOS ONE | https://doi.org/10.1371/journal.pone.0247331 February 19, 2021 8 / 18