Research Projects

Below are only projects that I participated at Virginia Tech. Please see my CV for a full list.

Current Research

Mood state in transport environments (Dissertation project)

A A

This study explores the relationship between satisfaction, affinity (a.k.a. the positive utility of travel), and travel demand and well-being. I applied the hybrid choice model, an extension of the traditional discrete choice, to incorporate the psychological factors into mode choice models.

The data were collected in Fall 2016-Spring 2017 (dataset 1, 2-week tracking) in Blacksburg-Roanoke, Washington DC, and Minneapolis, and Fall 2017-Spring 2018 (dataset 2, 1-week tracking) in Blacksburg-Roanoke and Washington DC.

Faculty advisor/PI: Steve Hankey (Virginia Tech). Funded by Virginia Tech Institute for Society, Culture, and Environment; VT College of Architecture and Urban Studies; and multiple sources.

[Link to study website]

Towards real-time air quality models

This project aims to add temporal precision to empirical models of air quality. My role is to process GPS data and merge different real-time datasets for exposure assessment.

Supervisors/PIs: Steve Hankey (Virginia Tech). Funded by Virginia Tech Institute for Critical Technology and Applied Science.

National scale direct demand model of pedestrian and bicycle traffic

A

This project employs NBPDP data and other count data obtained from local agencies to model peak-period bicycle and pedestrian traffic at the national level. This will enhance the generalizability and transferability of the existing direct-demand models. The final models would aid communities with few or no non-motorized traffic counts in selecting suitable sites to invest bicycle and pedestrian infrastructure, as well as quantifying and predicting pollution exposure and crashes.

Supervisors/PIs: Steve Hankey & Ralph Buehler (Virginia Tech). Funded by MATS-UTC through the USDOT University Transportation Centers Program.

Women and Cycling

A A A

This study explores women’s cycling behavior through a national survey on women conducted by the Association of Pedestrian and Bicycle Professionals (APBP).

[Link to study website] and [Exploratory Analyses]

Street Noise Relationship to Vulnerable Road User Safety

A

This study develops a method for evaluating street noise and documented crash rates for roadways in Austin, Texas, and Washington, D.C. using crowdsourced data collected from a smartphone app.

I developed bicycle models to predict bike traffic, which will become inputs to the street noise models.

PI: Greg Griffin (Texas A&M University). Funded by Safe-D UTC through the USDOT University Transportation Centers Program.

[Link to study website]

Past Projects

Spatial analysis of non-motorized crash on the street network: A case study in Northern Virginia

A

This study aims explore a regional scale of crashes involving pedestrian and bicycle. It provides insight on how crash locations are clustered on the spatial dimension, and how it affects other non-automobile users such as transit riders who bike or walk to access transit stops. The study helps promote active transportation and car-independence lifestyle as safety is among the biggest factors for people to choose to bike and walk.

Using Spatiotemporal Transit Accessibility to Predict the Transit Use of Residents in the Catchment Area (Master’s Major Research Paper)

This study proposes a way to measure spatiotemporal accessibility, as well as examines its impact, along with other transit attributes and built environment characteristics, on transit use of residents who live within the transit catchment area. We used the General Transit Feed Specification (GTFS) data, the Longitudinal Employment and Household Data 2011, and the longitudinal Neighborhood Activity and Travel Survey (N = 1,757) from the neighborhoods along the newly-opened Exposition light rail line in Los Angeles, CA from 2011 to 2013. We employed network analysis to measure job accessibility by transit, then built four sets of logistic regression models to assess the likelihood of using transit.