Sampling Bus Ridership at the Route Level: Initial Results from Efforts to Improve the Precision of Sample Estimates

UMTA's Section 15 reporting requirements establish precision standards and corresponding sampling plans for estimating bus ridership at the system level. However, many transit providers are interested in recovering data with sufficient precision to permit ridership analysis at the route level. One outcome of extending a system sampling plan to route level data collection and analysis would be a large increase in the sample sizes required to achieve a reasonable standard of precision. Given the costs involved in expanding the data recovery process beyond what is required for Section 15 reporting, alternative means of improving the precision of route level data warrant consideration. This report presents results from an initial effort to estimate ridership per bus trip on the basis of route level characteristics (primarily related to time of service). The analysis covers 17 routes served by the Tri-County Metropolitan Transportation District of Oregon (Tri-Met). We find that variations in ridership across bus trips within each of the routes studied are significantly related to the route characteristics. The effects of these characteristics on ridership were also found to vary from route to route. Improvements in the precision of ridership estimates attributable to the contribution of route-specific information translate, in the limit, to a 45 per cent reduction in the sample size required to achieve a given level of precision. Thus the potential gains in sampling precision from further research on this subject appear promising.