Graphical information processing of quantitative data

This dissertation presents a method for mapping large amounts of quantitative data, organized into tables, into graphical presentations. The presentations are called interactive matrix charts. Each chart consists of a grid containing graphical marks in each of the cells, and these marks may vary in color, size, shape and other characteristics. The goal of the interactive matrix chart is to aid researchers in the process of analyzing data by expressing the entire data set as a massive graph and, therefore, visually exposing underlying patterns.;The charts may be manipulated through both a simplification and a query mechanism. Simplifying the interactive matrix chart involves rearranging rows and columns of the chart. Querying the interactive matrix chart involves redesigning the individual marks meant to represent the underlying data. By simplifying and querying the interactive matrix chart, different relationships among the data are revealed.;This dissertation also describes and evaluates computer software that embodies these principles. The evaluation sought to determine whether or not medical science researchers who worked with quantitative data organized into tables were able to recognize overall patterns existing in their data. The evaluation concluded that researchers will indeed begin to see overall patterns in their data. However, the researchers experienced a notable paradigm shift from a more classical approach of examining micro-relationships between one or two data points to a novel approach of examining overall relationships among all of the data. It is concluded that the interactive matrix chart is a viable tool in the human-computer cooperative research environment.