STATCOM at the University of Washington: Implementing a Peer Evaluation System for a Local Fire Department

• Activity Number: 113
• Type: Contributed
• Date/Time: Monday, August 4, 2008 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistical Consulting
• Author(s): David W. Lockhart*+ and Julian Wolfson and Annette Ghee and Rori Rohlfs
• Companies: University of Washington and University of Washington and University of Washington and University of Washington
• Keywords: STATCOM network ; statistical consulting ; pro bono statistics ; survey methods ; missing data ; mixed effects model

Abstract

Statistics in the Community (StatCom) at the University of Washington (UW) was founded in 2005. One of the largest and longest-running projects we have participated in so far involves a local fire department serving a community of approximately 50,000 north of Seattle. After some initial experience developing a peer evaluation system, fire department representatives approached UW StatCom with questions about how to best analyze these data and assess potential biasing effects. Ultimately, we became a partner in refining the entire peer evaluation procedure. We have provided support on questionnaire design including acceptability of the changes to the system, sampling to select raters, and data analysis. In this poster, we provide an overview of UW StatCom activities and describe some of the more interesting features of the fire department project.