Multiattribute Data Presentation and Human Judgment: A Cognitive Fit Perspective*

Abstract We assessed the ability of the cognitive fit theory to explain the performance of certain display formats on multiattribute judgment tasks. This theory suggests that for most effective and efficient problem solving to occur, the problem representation and any tools or aids employed should all support the strategies (methods or processes) required to perform that task. The theory was tested by assessing performance with schematic faces, graphs, and tables on a bankruptcy prediction task. Bankruptcy prediction involves integrating a large amount of data (a number of financial indicators over a number of years), as well as referring to ranges and/or levels of financial indicators. Schematic faces provide a cognitive fit with such tasks since the information in a face can be processed holistically; however, they do not permit decision makers to refer to the underlying data. Graphs facilitate a different integrating process; further, they preserve characteristics of the underlying data. Tables, on the other hand, do not aid the decision maker in integrating information; they provide only the underlying data values. It was hypothesized that graphs would provide the best cognitive fit for the bankruptcy prediction task since they permit processing both integrated and discrete data. Participants made judgments with two of the three display formats, at two levels of information load, in a fractional factorial design. The information load manipulation was designed to provide meaningful and meaningful plus redundant information to the decision maker in a test of information load “per se.” The research findings provided substantial support for the theory of cognitive fit. The findings also have implications for the study of information load.