With the knowledge that Texas had purchased a large add-on to the 2009 National Household Travel Survey (NHTS), the North Central Texas Council of Governments (NCTCOG) Travel Model Development group initiated an in-depth study of the NHTS dataset with the main goal of determining how it can be used to update the region’s travel demand model. The process involved investigation of the sampling and data collection methodologies, cleanup of weekday samples, re-expansion to 2011 ACS data, and comparisons with other surveys both temporally and geographically.
One important finding was revealed when trying to complete the 220 records which were missing household income. Testing the use of the hot deck imputation on sample records found that the method was not satisfactory and a complex expansion could produce better results.
The updated expansion process enhanced the original NHTS method through stratification of samples by weekday, socio-economic characteristics and geographical consideration along with use of the 5-year ACS 2011, the closest ACS year to the NHTS dataset that benefits from Census 2010. The new expansion process uses a multi-dimensional waterfall method on household and person records with its own advantages and dis-advantages.
NCTCOG’s history of household surveys provided an opportunity to compare some of the 2009 NCTCOG NHTS results with local surveys performed in1965, 1984, and 1996. Among the comparisons made with previous surveys, understanding home-based work trips differences introduced a challenge. It was also possible to compare the results with the NHTS results across the nation. This presentation will provide a summary of the process followed by NCTCOG, and show several key statistics on trip rates, trip length, trip purpose, and socio-economic characteristics for both temporal and geographical dimensions.