This proposal is for two one-week workshops for 20 social scientists, each covering techniques of exploratory data analysis using a statistical package particularly suited for graphical data analysis, Stata. Participants will develop skills in using and teaching robust statistics, analytical graphics, and data transformations. Exploratory data analytic techniques, such as data smoothing, median polishing, re-scaling, and transforming data using the ladder of powers, five number summaries, box and whisker plots, and stem and leaf charts, will be emphasized. The workshop will also illustrate pedagogical methods for teaching EDA techniques at the undergraduate level. The significance of the project to undergraduate faculty and students lies in the use of exploratory data analysis techniques as interesting and useful alternatives to traditional confirmatory statistics. Faculty who have training through the level of multiple regression will be accepted for one week, while a two-day longer workshop will accept faculty whose training is up to the multiple regression level. In both cases, faculty will have responsibility for the required methodological courses in their disciplines.
