Starting Points Matter: Spatial Variation in Marginal Effects for Negative Binomial Trip Models
Open Access
- 9 May 2022
- journal article
- research article
- Published by Network Design Lab - Transport Findings in Findings
- p. 35459
Abstract
Several publications studying associations between land use and shared micromobility (such as dockless scooters and bicycles) rely on Negative Binomial regression models or similar, reporting the models’ untransformed coefficients. We demonstrate a new way of reporting the associations identified by such models: By reporting marginal effects for several different starting points resembling real-world locations rather than just coefficients, these models’ implications can be made more approachable to a lay audience. At the same time, we draw attention to the models’ limitations when applied to locations that are outliers in terms of density.Keywords
This publication has 5 references indexed in Scilit:
- Spatial interactions of shared e-scooter trip generation and vulnerable road user crash frequencyJournal of Transportation Safety & Security, 2021
- Influence of the built environment on E-scooter sharing ridership: A tale of five citiesJournal of Transport Geography, 2021
- A segment-level model of shared, electric scooter origins and destinationsTransportation Research Part D: Transport and Environment, 2021
- Dockless E-scooter usage patterns and urban built Environments: A comparison study of Austin, TX, and Minneapolis, MNTravel Behaviour and Society, 2020
- Understanding the Shared E-scooter Travels in Austin, TXISPRS International Journal of Geo-Information, 2020