For estimating the system total unlinked passenger trips and passenger miles of a fixed~route bus system for the National Transit Database (NTD ), the sampling plans approved by the Federal Transit Administration (FTA) may either oversample or do not yield FTP:s required confidence and precision levels for the specific conditions of a transit agency. This guide helps transit agencies avoid these problems by developing sampling plans customized to their fixed-route bus services. Detailed steps are provided to calculate the statistical variation in passenger miles and unlinked passenger trips and the correlation between them. The sampling plans that transit agencies develop using this guide maintain the main features of the original FTA plans. More important, transit agencies can easily use this guide to develop customized sampling plans in an Excel environment using three existing pieces of information: an NTD sample that they have collected with their current FTA approved sampling plan; annual days of service; and daily total number of one~way bus trips. This guide complements FTA Circular C2710.1A in that transit agencies first select one of their own sampling plans that best meets their staffing needs, and then follow the procedures in the circular on sampling and collecting field data. Transit agencies should be able to use this guide and the related calculations as an approval of a qualified statistician as required by the current NTD reporting manuals.
Chu, Xuehao & Ubaka, Ike.
A Guide to Customized Sampling Plans for National Transit Database Reporting.
Journal of Public Transportation, 7 (4): 21-47.
Available at: http://scholarcommons.usf.edu/jpt/vol7/iss4/2