Smith, Benjamin KeniraySmall multiples play a vital and growing role in the display of complex information. They are particularly useful for depicting spatiotemporal data, for which more traditional graphs and maps are inadequate. However, the scientific investigation of the usefulness of small multiples has been limited and often misdirected. In five experiments, small bar graphs are used to investigate several factors that could influence the comparability of the small graphs that comprise a small multiples graph. These factors include the distance between the graphs, the alignment of the graphs, the orientation of the bars, the length of the bars, and whether the graphs contain a single bar or multiple bars. In all cases, the most important factor affecting the comparability of the graphs was the difference in lengths, or difference in the increase of lengths, that the participants were asked to compare. The effects of distance were greater when the bars were closer to each other than when they were farther apart, suggesting that the bars are Title of Dissertation: COMPARING SMALL GRAPHS: HOW DISTANCE, ORIENTATION, AND ALIGNMENT AFFECT THE COMPARABILITY OF SMALL MULTIPLE BAR GRAPHS Benjamin Keniray Smith, Doctor of Philosophy, 2012 Dissertation Directed By: Professor Kent L. Norman, Department of Psychology compared using central vision. For pairs of graphs with a single bar each, comparability decreased as the distance between the graphs increased, although this effect was more prominent measured by accuracy than response time. Graph arrangements with horizontal alignments and vertical orientations were more comparable, although these effects were more subtle than the distance effects. For pairs of graphs with two bars each, the distance between the graphs had no effect on the accuracy of the comparison, and only a slight effect on the response time. Alignment and orientation had no effect on the comparability of graphs with two lines. The similarity of the lines in each graph, including but not limited to the critical length increase, significantly affected the comparability of the graphs. Part of a graph difficulty principle for small multiple graphs is offered as advice for graph creators.DissertationCognitive psychologyComputer sciencebar graphhuman-computer interactioninformation visualizationsmall multiples