What Is a Transportation Statistician?

Li LeungLi Leung serves as a survey statistician within the Office of Survey Programs at the Bureau of Transportation Statistics. She is a key team member of the Commodity Flow Survey and National Census of Ferry Operators. Leung earned her MS in civil and environmental engineering with an emphasis in transportation engineering and a BS in applied computational mathematical sciences with an emphasis in statistics from the University of Washington.
 

Have you ever heard that it is safer to fly than to drive? That factoid—based on data and statistics from the National Highway Traffic Safety Administration and National Transportation Safety Board—is provided to you by a transportation statistician.

The wonderful thing about transportation is that it touches almost everything we do—how we get to work, how goods get to market, and how we travel for fun. A lot of information is collected for some of the transportation modes, such as aviation, highway, marine, mass transit, pipeline, and rail. Other modes are inspiring transportation statisticians to think of new ways to measure movement such as using smart phones to measure pedestrian activity or GPS devices to track bicycle trips.

A transportation statistician can be involved in devising new ways to capture information and may use some of the data available to measure or project safety risks and fatalities, improve logistics for freight delivery, or help develop plans and policies that can change the way the world moves. Let’s highlight four professionals who apply their statistical skills to transportation.

 
Linda Ng Boyle

Linda Ng Boyle is an associate professor at the University of Washington. She directs the Human Factors and Statistical Modeling Laboratory, which emphasizes the use of appropriate analytical tools to solve problems related to human factors and transportation.

 

Linda and her researchers examine the influence of driver behavior on crashes, injuries, and other unsafe events. This research relies heavily on statistics to quantify the complex nature of drivers. Driver behavior data can be collected in near real-time, retrospectively, in controlled laboratory settings, and longitudinally. The combination of data sources provides a comprehensive picture of issues related to driver safety, but requires appropriate statistical techniques to assimilate different sources.

“Some of my current projects include quantifying the impacts of distractions, fatigue, and high workload on teenage drivers, commercial drivers, or those with cognitive impairments,” explained Linda. She says that differences in drivers include many factors and systems designed for safety, but which may have unintended consequences with prolonged use. Statistics provide the opportunity to examine these factors and provide insights for automobile designers, policymakers, and education.

Looking into the future, Linda sees that data collected on a more connected environment with the driver, vehicle, and road—often called vehicle infrastructure integration—would provide statisticians with data sets that can help home in on unsafe driving situations. Some of the challenges include the exponential explosion in the data available, which could make correlations and trends more obvious. However, this fine-grained data may be quite noisy and require new tools for data abstraction to avoid misleading outcomes.

 

William C. Davie Jr. is the assistant division chief for research and methodology for the Service Sector Statistics Division at the U.S. Census Bureau. He has worked at the Census Bureau for nearly 20 years and on several surveys related to transportation industries such as the Service Annual Survey (SAS), Quarterly Services Survey (QSS), and Commodity Flow Survey (CFS).

 

The SAS and QSS provide national estimates of revenue and expenses for many service industries, including transportation of passengers and cargo, warehousing and storage of goods, scenic and sightseeing transportation, and support activities such as freight transportation arrangement. The CFS, conducted every five years as part of the economic census and in partnership with the Department of Transportation’s Bureau of Transportation Statistics, measures the value and weight of shipments by manufacturing, mining, wholesale, select retail and service, and auxiliary establishments by mode, commodity, origin, and destination at the national and subnational geographic levels.

To conduct these surveys, statistics are used to design and select representative samples; determine methods to edit data and adjust for nonresponse; develop weighting, estimation, and variance estimation methods; apply disclosure avoidance techniques; and ensure data products meet Census Bureau quality standards. Each survey presents different statistical problems.

Bill said, “I like my job because there is always something new to explore, whether it is to expand the SAS and QSS to include more industries or developing and testing the use of Internet response for the 2012 CFS. The mathematical statistician at the Census Bureau must continue to research new ways of collecting, processing, and disseminating the survey results as the demand increases for more detailed, timely, and useful measures of the transportation industry.”

 

Tanya Rodríguez is an analyst at Abt SRBI’s Transportation and Regional Planning Practice. She recently completed a master’s degree in urban planning; now her job marries this field with survey research. She has been responsible for numerous transportation projects, including customer satisfaction, service evaluations, household travel diaries, and commercial trucking practices. These studies involve sampling and inferential/multivariate statistics. For example, on a commuter rail customer satisfaction survey, she implemented the sampling strategy to selected trains/cars to yield a representative sample of passengers. She also prepared the final report that, along with other data, was used to determine resource allocation for future capital improvements. These types of reports use a variety of inferential and multivariate techniques.

 

Tanya has always wanted to work in the public realm and contribute to policy decisions that improve communities and urban life. This job has given her the opportunity to do just that. She enjoys interacting with transit and other government agencies on understanding the needs of different transportation markets. She is particularly interested in making sure under-represented populations are not misrepresented or completely omitted in policy decisions.

Tanya said, “The survey industry is changing rapidly—we are working with computers, smartphones, and GPS units to collect public opinion and travel data. Twenty and 30 years from now, transportation surveys will be using even more unique and innovative ways of using less financial capital while still ‘getting the job done’ without sacrificing quality.” She is looking forward to exploring new survey methods that incorporate nonobtrusive digital measurement and the traditional survey approach using personal interviews that rely on recall.

 

Tom Krenzke is an associate director in the statistical group at Westat. He has developed expertise in methods for statistical disclosure control (SDC)—an important and evolving area—through applications involving transportation data. In a highly collaborative effort that included transportation planners and U.S. Census Bureau staff, Tom led tasks to apply SDC treatments to American Community Survey (ACS) data to reduce disclosure risk to a level that met standards for disseminating the Census Transportation Planning Products while limiting the effect on data utility. The SDC approaches maximize the amount of data to be released to the public. Tom likes to work on transportation data to help ensure quality data is released and found it rewarding to work collaboratively with several teams.

 

Tom sees the amount of transportation data escalating through automated data collection approaches and the use of GPS devices for transportation studies. With so much data, he believes “graphical data analyses, which provide a clear and concise message, are going to be important tools for the statistician in the near future.” However, statistical standards for accounting for propagated error in the reporting of precision estimates (i.e., the margin of error) will be necessary.

Another area for the near future is the development and maintenance of centralized online research data centers that allow transportation data to remain safe from disclosure concerns while research statisticians explore them as they relate to our health and our need to strive for new efficient use of energy.

Being a transportation statistician offers opportunities to promote many people’s lifestyles and movement of goods and services, thus playing an indirect, but key, role in society. Statistics on all aspects of transportation are important, as they are used to help determine how to allocate funds to improve the roads, bridges, waterways, railways, and airports used around the clock to transport people and goods throughout the United States. Both transportation and statistics are interdisciplinary; hence they intertwine with urban planning, economics, business, engineering, computer science, politics, and research. There are opportunities in the private and public sector, as well as in academia.

Linda, Bill, Tanya, and Tom have careers that mix various disciplines. A career as a transportation statistician can be rewarding, with challenges that incorporate new policy objectives, foster new discoveries, and generate new knowledge extension. The transportation statistician provides reliable data-driven information to decisionmakers and can make a difference in improving our world.

Print Friendly, PDF & Email