Conditioned Micromaps Based on Status and Trend Confidence Intervals

• Activity Number: 543
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Graphics
• Author(s): Daniel Carr*+ and Krista Heim
• Companies: George Mason University and George Mason University
• Keywords: Multivariate maps ; statistiscal stories ; confidence intervals ; simple appearance

Abstract

Micromaps designs provide several ways to show or use estimate uncertainty in a spatial context. Linked micromaps from http://statecancerprofiles.cancer.gov/micromaps/ routinely include confidence intervals. This talk addresses a variation on conditioned micromaps. For regions such as US states the design represents three-level class memberships for each three variables using a 2-way a grid of micromaps and color. Rather than using the dynamic software, CCmaps, with its three three-class dynamic sliders to partition states, the examples use confidence intervals to partition states into status and/or trend classes. One example is based on standardized test averages for 4th grade and 8th grade mathematics and 8th grade reading. This shows a strong (below, below, below) and (above, above, above)pattern. Only Illinois is similar to national average for all three tests. Another example shows that almost all states had rising liver and bile duct cancer mortality rates. CDC found this to be associated with hepatitis C in baby boomers. There are some statistical issues but the design has simple appearance and should be very useful for telling multivariate statistical stories.