Statistical Analysis for Motor Vehicle Crash Insights

The National Center for Statistics and Analysis (NCSA) is responsible for providing a wide range of analytical and statistical support to NHTSA and the highway safety community at large. NCSA collects data on police-reported motor vehicle traffic crashes in the U.S. through its various data collection systems. NCSA needs advanced statistical and data tool support to provide more complete and analysis-ready data, improve and expedite data analytic processes, develop alternative models for analysis, develop new data tools and resources to share insights gained from the data in a timely fashion. This includes applying advanced statistical algorithms to enhance current and additional data elements, develop data imputation techniques to handle missing data, integrate supplemental data resources to augment the data, develop alternative statistical crash projection and forecasting algorithms. This also includes creating data query tools, user-friendly analysis applications, and dynamic dashboards for quicker access to insights and trends. NCSA also needs support in investigating use cases for emerging techniques, identifying supplemental data sources, developing statistical surveillance programs for special studies, and creating both technical and non-technical training and documentation for end users. NHTSA’s Office of Traffic Records and Analysis (OTRA) Mathematical Analysis Division (MAD) is soliciting the help of a contractor with following experience and expertise:

• Demonstrated experience in conducting gap analysis, including identifying ways to improve research, data, and analysis, consolidating findings, and developing statistical solutions to address the identified gaps and needs
• Demonstrated experience handling missing data and expertise in developing custom data imputation algorithms using statistical techniques customized for the imputation at hand
• Demonstrated experience identifying supplemental data sources such as roadway infrastructure data and exposure data for behavioral research topics (data related to distraction, speeding, micro-mobility usage)
• Demonstrated experience in applying advanced statistical algorithms for estimation and forecasting (e.g., ARIMA models, seasonal adjustment, time series cross-section regression, naïve Bayes, and gradient-boosted models) 
• Expertise in considering statistical nuances such as changes in variance estimates, additional bias components, changes in inference due to transformations or other model or data adjustments
• Expertise in developing and customizing off-the-shelf advanced statistical and emerging computational algorithms. Knowledge in modifying variance components, changes in result interpretation, bias effects, and accuracy when altering advanced models
• Expertise in geospatial analysis techniques (e.g., kriging, geo-weighted regression, spatial neural networks) and geospatial data integration
• Proven experience developing data visualization products using dashboard software, developing data query tools, and working with large data systems, developing statistical programs using common statistical software

• • The contractor shall perform an in-depth review of the data, tools, publications, analytical products, analysis, and documentation specified by the COR(TO). This includes reviewing NCSA data, NCSA tools, publications, platforms (CDAN, EDMAS), NCSA documentation, and analysis conducted by NCSA. This also includes reviewing data sources used for exposure metrics such as Population data from Census Bureau, Vehicle Miles Traveled (VMT) data from Federal Highway Administration (FHWA), the National Performance Management Research Data Set (NPMRDS) from FHWA, the Registered Vehicles data from FHWA and R.L. Polk & CO, the National Household Travel Survey (NHTS) from FHWA, and the Web-based Injury Statistics Query and Reporting System (WISQARS) from the Center for Disease Control (CDC). Publicly available state data analysis products and tools shall also be reviewed. The contractor shall provide a briefing and report highlighting key findings from this review.

• TBD
• District of Columbia