Text

theses

Gathering and integrating information from spatially distinct and qualitatively different sources requires decisions about when and where to sample information. An example is reading graphs and accompanying text to arrive at a coherent interpretation. Eye movements were recorded while subjects viewed bar graphs depicting information about two attributes of two fictitious products, along with descriptive text, after which subjects indicated which product they preferred. Three perceptual-motor configurations were tested: (1) Simultaneous: Graph and text were displayed adjacent to each other; (2) Button-Press: Graph and text appeared sequentially, with the appearance of each triggered by a button press; and (3) Eye-Contingent: Graph and text appeared sequentially, in the same spatial locations as in the Simultaneous condition, with the appearance of each triggered by a saccade into the region. Shifts of gaze between graph and text occurred about twice as often in the Simultaneous condition than in either Button-Press or Eye-Contingent conditions. The rate of shifts in the Button-Press and Eye-Contingent conditions did not differ, showing that the relevant factor was not motor effort (saccade vs. button press), but rather the sequential vs. simultaneous aspects of the presentations. Conditions did not differ either in trial duration or in the proportion of time spent in the graph or text. In the Simultaneous condition, most trials began with relatively long inspections of the graph and the text, with a preference to view the text first. The initial inspections of the graph or text were then followed by one or two relatively brief visits to each. Patterns were different for the other two perceptual-motor conditions in that the initial inspections of graph or text were typically longer and the subsequent visits to each were rare. Analyses of fixated locations showed a preference to use the more frequent visits in the Simultaneous condition to re-examine previously seen material rather than to look at new material. The simultaneous availability of graph and text did not necessarily encourage a strategy of inspecting each region in segments but spread out the reviewing time across subsequent visits to graph and text. These results show that the strategies used to integrate information across graphs and text depend on more than just the information content, and are influenced by even relatively minor variations of perceptual availability. The extra steps involved in conditions aside from the Simultaneous condition may discourage saccades as a part of an overall strategy of conserving neural costs or cognitive load. The relative perceptual availability of information in different spatial regions must be taken into account when developing models of cognitive strategies on the basis of observed saccadic patterns.

ETD_8326

M.S.

Includes bibliographical references

by Jason F. Rubinstein

doi:10.7282/T3SF309W

ETD graduate

The author owns the copyright to this work.