Ordinal Response Variables Observed Continuously Over Time: Data from Shadow, Gray Eagle, and OSRVT

• Activity Number: 370
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Defense and National Security
• Author(s): Matthew Avery*
• Companies: Institute for Defense Analyses
• Keywords: Functional Data ; Ordinal Response ; Defense

Abstract

Recent operational tests for the Shadow and Gray Eagle (two unmanned air vehicle (UAV) systems employed by the US Army) have used ordinal response metrics to assess the system's ability to provide a continuous video stream to the One System Remote Video Terminal (OSRVT), the system employed by the end user to view data from these UAVs. Data collectors assign levels from an ordinal scale continuously over the duration of each mission while covariates such as the antennae used by the receiving unit and the distance of the broadcasting air vehicle to the ground receiver are also recorded. Thus, we are faced with a problem of an ordinal response variable observed continuously over time. Since the covariates are also observed continuously over time, we are faced with a problem best suited to the tool of functional data analysis. However, the problem of ordinal responses observed continuously over time is not well discussed in the literature. Here, we present a variety of approaches for analyzing this unique but relevant type of data.