Extracting information from graphics

Abstract

Graphics are used in many situations whae a large set of data is to be presented in an effective manna. Having a set of standards for the constructicxi of graphics is critical for infcxmatioi graphics. The current work is the first step in a project motivated to establish connections between applicati «i-criented work (eg, Tufle, 1983) and the theoretical work (eg, Qeveland & McGill, 1984; Pinka, 1990). Our model is an integration of Pinker's (1990) and Qeveland and McGill's (1984) theories. Our model of graphics comprehension states that comprehension of information represented in grafiucs is affected by the organizational tendencies and the UmitaticHis of the paceptual system, capacity ccMistraints of the wcM-king memory, and knowledge stored in the memory system. AccOTding to our model, there are a number of reasons why a graprfi reader may have difficulty. First, the graph may require the use of inferential and tqj-down processes. Seccmd, the schema fw a type of graph may not contain an indicaticm fcff needed infcMrnation. Third, just as in any other cognitive task, the capacity of processing resources may limit the amount of infcrmation that can be manipulated at once. Finally, the graprfi schema may mislead the informaticxi search because the search relies primarily on defeult patterns.