A Systematic Review on the Practice of Evaluating Visualization

User studies in cartography: opportunities for empirical research on interactive maps and visualizations

“…As with cartography, there is increasing emphasis on user-centered evaluation in the related fields of information visualization and scientific visualization as the essential mechanism for determining whether an interactive design 'works' (e.g. Barkhuus & Rode, 2007;Borkin et al, 2011;Isenberg, Isenberg, Chen, Sedlmair, & Möller, 2013;Lam, Bertini, Isenberg, Plaisant, & Carpendale, 2012). But, which empirical methods should we use, and at what times during design?…”

An overview of semantic search evaluation initiatives

“…Tasks could include: specific fact-finding (e.g., finding someone's phone number), extended fact-finding (e.g., finding books by the same author), open-ended browsing (e.g., identifying any new work on voice recognition from Japan), and exploration (e.g., finding out about your family history from an online archive) (Shneiderman, 1986, p. 512). For more browsingoriented tasks, such as exploration, then evaluation may go beyond simply dealing with ranked lists of results and assessing visualisations (Isenberg et al, 2013).…”

Stacked space-time densities: a geovisualisation approach to explore dynamics of space use over time

“…This is not applicable in our case, since we use the same rendering for both volumes. Instead, we adopt two frequently applied evaluation methods from information visualisation [29], algorithm performance (left green arrow) and qualitative inspection (right green arrow) Instead, we employ evaluation measures from information visualisation [29] and use the two most prevalent evaluation methodologies in the visualisation community: Qualitative Result Inspection (QRI) and Algorithm Performance (AP). QRI is a method where a certain visualisation methodology is used to produce an image that demonstrates how something that could not be seen before can now be seen with the new method.…”

“…QRI is a method where a certain visualisation methodology is used to produce an image that demonstrates how something that could not be seen before can now be seen with the new method. An extensive review of information visualisation studies [29] found that QRI was used in 95 % of all cases. An example of this methodology used for 2D trajectory density can be found in [49].…”

“…This means that they are indistinguishable if considered only in 2D space and if only 2D density distribution is used to visualise the home range. We also evaluate the stacked densities algorithm against the geometrically optimal space-time density [13] with regards to identifiability of crucial visual elements (using Qualitative Result Inspection, QRI [29]) and computational efficiency (using Algorithm Performance measures [29]). We further apply the visualisation method to a case study of trajectories of lesser black backed gulls (Larus fuscus) to identify two different types of bird behaviour by distinguishing between spatio-temporal and spatial-only hot spot, similar to those in our synthetic example.…”