Collaborative Research: Quality Graphics for Federal Statistical Summaries
Supported by the National Science Foundation through their Digital Government program.

The Federal government distributes a vast quantity of statistical summaries in printed and electronic form. The full wealth of information that might be derived from these summaries is not being realized because limited attention is paid to disseminating summaries in understandable forms.

The objective of this research is to develop and assess quality graphics for federal statistical summaries. The development and assessment process will consider perceptual and cognitive factors in reading, interacting with, and interpreting statistical graphs, maps, and metadata representations. The purposes of the quality graphics include exploration by agency users evaluating data quality and looking for emergent trends, decision making by public policy makers, and communication of statistical summaries to the public.

The proposed research addresses four topic areas: converting tables to graphs and maps, representing metadata, interacting with both graphs and maps, and conveying multivariate spatial and temporal relationships. The research features use of Web-based “middleware” components to provide rapid development of graphics for usability testing.

The featured middleware is a Java Graphics Component Library (GPL). A long history of research has recently culminated in a rigorous graph algebra that is the foundation for the library. The library also benefits from the collective intellectual effort being poured into Java. Principles of human perception and cognition will be used to guide the construction of statistical graphics, and usability tests will help to refine those principles in the context of federal statistical summaries.

The research team has been carefully selected to bring together outstanding and compatible investigators. This group provides expertise in statistical graphics, cartography, human perception and cognition, and software development. The team is composed of university, industrial, and federal agency researchers that have the collective expertise needed to conduct the research. For example, one industrial partner, AT&T, has addressed scalability issues and has been dealing with five gigabytes of data a day, every day. The collaborators from agencies endorsing this research (representing BLS, NCI, NCHS, EIA, NASS, CENSUS, EPA, and BTS) provide the domain expertise and vision to develop software products that will be useful to the agencies.

Research products include a better understanding of principles of cognition and perception as applied to static and interactive statistical graphs and maps, improved middleware components, and evaluated demonstration software than can lead to production software for web-distribution of statistical summaries. The graphics software will embed elements fostering statistical literacy. Beyond the usual papers and conference presentations, the milestones include annual multi-agency presentations in Washington, D.C., to promote transfer of methodology.

The research promotes communication advances sought by the President’s committee on information technology. The results will impact the public through better communication, understanding, and policy in the areas of health, environment, energy, agriculture, employment and transportation.

This material is based upon work supported by the National Science Foundation under Grant No. 9983451, 9983459, 9983461.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

all material copyright © 2000-2002
web site hosted by the PSU GeoVISTA Center