Kypridemou, ElektraZito, MicheleBertamini, MarcoKerren, Andreas and Garth, Christoph and Marai, G. Elisabeta2020-05-242020-05-242020978-3-03868-106-9https://doi.org/10.2312/evs.20201039https://diglib.eg.org:443/handle/10.2312/evs20201039The way in which a graph is described visually is crucial for the understanding and analysis of its structure. In this study we explore how different drawing layouts affect our perception of the graph's properties. We study the perception of connectedness, tree-ness and density using four different layouts: the Circular, Grid, Planar and Spring layouts. Results show that some layouts are better than others when we need to decide whether a graph is a tree or is connected. More sophisticated algorithms, like Planar and Spring, facilitate our perception, while Circular and Grid layouts lead to performance not better than chance. However, when perceiving the density of a graph, no layout was found to be better than the others.Attribution 4.0 International LicenseHuman centered computingGraph drawingsEmpirical studies in visualizationComputing methodologiesPerception10.2312/evs.202010391-5