Sec. 975.7. Statistical sampling provisions

(a) In order to ensure the submission of valid revenue passenger statistics in all instances where the system submits “estimated” data, the STOA program will impose, effective January 1, 1990, annual precision (10%) and confidence (95%) requirements on any system's sampling and/or data collection technique used to calculate estimated passenger data and any revenue passenger-related data. Thus, any system that does not submit actual revenue ridership data must be able to document that its annual passenger statistics submitted for STOA from April 1990 on, will be in compliance with these annual confidence and precision limits.

(b) For demand responsive systems, however, it is required that such operations document and present “actual” quarterly revenue passenger statistics.

(c) Each public transportation system is required to categorize all riders into eligible revenue passengers (including cash fares, tokens and other pre-payment instruments that satisfy the per-passenger minimums specified in section 975.2 (d)) and into ineligible non-revenue passengers. It will be the responsibility of the public transportation system to certify that its sampling technique and sample size provide estimates of quarterly revenue passengers, such that when considered on an annual basis they satisfy the required levels of confidence and precision (95% and 10%, respectively).

(d) Vehicle Mile Data - Each public transportation system is required to calculate eligible revenue vehicle mile data by employing one of the two following methods:

(1) compute total actual miles and adjust for deadhead and other non-revenue miles, or

(2) derive revenue vehicle mile portion of total miles from schedules and adjust for missing trips; and

(3) furthermore, the transit system must certify and be able to demonstrate that there is a system in place to account for missing trips and/or deadhead and other non-revenue mileages. The system must also certify and be able to document that any standard mileages that are preassigned for individual routes reflect audited, actual mileages for such routes accurate to the nearest one tenth of a mile.