The Cleveland GPS Household Travel Survey: Survey Design, Imputation of Trip Characteristics, and Secondary Uses of the Data

In 2011, the Ohio Department of Transportation (ODOT) and the Northeast Ohio Areawide Coordinating Agency (NOACA) spearheaded the second large-scale GPS-based household travel survey conducted in the United States. The survey covered the five counties in and around Cleveland, OH that comprise the planning area of NOACA. Of the 6,542 households recruited into the Cleveland GPS-based travel survey, 4,001 households were asked only to return their GPS devices after use (with no additional participation requirement). Of the remaining 2,541 recruited households, 1,940 were asked to complete a prompted recall (PR) survey based on recorded GPS data while the other 601 households (which were comprised entirely of household members over the age of 75) were provided with simple travel logs to use for recording and reporting travel. All GPS households were asked to use the GPS devices for a minimum of three consecutive days (with at least two weekdays), resulting in the collection of 108,441 GPS trips covering 708,557 miles. Regularly visited locations were collected by the survey research firm during the recruitment interview, and were leveraged along with recorded GPS traces and land use data to assign trip purpose, mode, and travel costs to GPS-Only households based on a probabilistic ‘best fit’. The algorithms were calibrated by PR participant responses. This paper starts with an in-depth overview of the survey design and concludes with a review of several spatial analyses conducted by NOACA transportation planners using the collected GPS dataset.