Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
97Â Â Data Methodology Travel Survey The travel survey discussed in Chapter 3: Micromobility Users and Utilization was derived from the Populus Groundtruth dataset, which consists of representative data on transportation decisions, including public transit use, vehicle ownership, and new mobility service adoption and utilization, ride hailing, carsharing, bikesharing, and e-scooter sharing. The data in this report focus on the scooter users surveyed and the decisions they made around micromobility use and transportation choices. The results summarized here are based on a representative sampling of the populations of 18 metro areas: Atlanta, Austin, Boston, Chicago, Denver, Houston, Knoxville, Los Angeles, Memphis, Nashville, New York City, Portland, San Antonio, San Diego, San Francisco, San Jose, Seattle, and Washington, D.C. For the purposes of this analysis, the researchers examined Populus data collected from mid-May to mid-October 2019 from over 15,000 individuals. Based on rigorous methods, key demographic variables (age, income, race, and gender) from this sample match those of the actual populations at the metro level. The 18 metro areas were grouped by common population and transit metrics gathered from U.S. Census data, which helped to condense the resultsâ reporting. The metrics used to group the metro areas included population, population density, housing unit density, gross domestic product per capita, and transit ridership per capita. Based on the grouping, most of the metro areas sampled fell into the low-density, low-transit-use cluster. After reviewing the cluster standard error, the researchers found that it was not advantageous to increase the number of groups as it increased the error. Micromobility and scooter adoption by region was associated with the home region of the person surveyed. For example, a person who resides in New York City who had tried using a shared scooter while visiting San Diego would be counted as a scooter user who would be included in the New York regional sample. Weighting by Frequency of Use The results presented in the How and Why People Use Scooters section of Chapter 3 are based on trip-weighted responses, with weights equivalent to estimated monthly scooter rider- ship based on reported frequency of scooter use. While over 1,500 survey respondents had used shared electric scooters, some users were regular, frequent riders (daily or almost daily), while others had not ridden a scooter in the previous 3 months and may have only tried using them once while traveling. This trip-weighted method of analysis better reflects the impacts of the use of shared electric scooters in accordance to their actual utilization. A P P E N D I X A
98 Transit and Micromobility Micromobility Data Populus used two sets of data on micromobility for the Micromobility Usage Patterns and Impacts section of Chapter 4: ⢠Trip data obtained from docked bikeshare systems, and ⢠Trip data from dockless shared-scooter operators (with their permission and the permission of the cities where they operate) The scooter data were not collected specifically for this project but were used in an aggre- gated form so that individual trips and sensitive information about individual users and scooter company operations were obscured. Docked bikeshare data were cleaned to only include the regions in which the researchers were interested and over the time frame needed for the data comparison. Scooter trip data were used in the transit analysis to look at trips starting and ending around transit locations. For the maps of scooter usage in Chapter 4, data were aggregated by location (i.e., no individual trips are identified) and across operators. Trip origins and destinations were aggregated in hex-style representations to create a grid of equal size areas. Each hex area has an edge-to-edge length of approximately 175 meters. For each region, 1 monthâs worth of data (from October 2019) were used. The darker areas are those with more trip origins or destinations. Actual counts are not shown and vary by region. The maps show the relative count quantiles specific to that region.