Introduction
A common question in exploratory research is: How do several contrasting groups dier?" Learning
about group dierences is a central problem in many domains. For example, the US Census Bureau
prepares many statistical briefs that compare groups such as the publication, The Earnings Ladder:
Who's at the Bottom? Who's at the Top?" which contrasts high and low income earners over the
years 1979 to 1992. They report such facts as: About 4 in 10 year-round, full-time workers aged
18 to 24 had low earnings in 1992, up 19 percentage points since 1979.