Data Collection

Crash Data Improvement Program (CDIP) Final Report

INTRODUCTION

The Crash Data Improvement Program (CDIP) provides States with a means to measure the quality of the information within their crash database, and provides recommendations for improvement. It can provide States with measures to address the six data quality attributes of timeliness, accuracy, completeness, uniformity/consistency, integration, and accessibility. Additionally, the CDIP helps familiarize the collectors, processors, maintainers, and users with the concepts of data quality and how high quality data helps to improve safety decisions.

This report covers the years 2010 through 2013. The CDIP began in 2008 and has been delivered in 18 States, 10 which are under the current contract between VHB, Inc. and the Federal Highway Administration (FHWA) Office of Safety. Figure 1 shows the eight States that were completed before the start of the current contract, the 10 states completed at the time this report was developed, and four more States that are scheduled for completion before the end of the contract in 2013. This will bring the total number of CDIP States to 22 by the end of 2013.

Figure 1: All CDIP States 2008-2013.

CDIP started in response to weaknesses in States' submissions to the Section 408 grant program (Traffic Records Improvement Funding) under the preceding Transportation Authorization Bill: SAFETEA-LU (Safe Accountable Flexible Efficient Transportation Equity Act–a Legacy for Users). The panel reviewing States' grant applications observed that many States had few (if any) quantitative measures of crash data. Many of the data quality measurements that States reported provided only project-specific views of data quality rather than the desired global, system-wide perspective that was called for in the legislation and the Section 408 grant application guidance. It was clear from the quality of the submissions, and from Traffic Records Assessments, that many States lacked a formal, comprehensive crash data quality management system. Moreover, it appeared that few States truly understood the methods and importance of data quality measurement. The CDIP has been intent on providing States with guidance on assessing and improving crash data quality. The delivery mechanism to States has used the expert advice of a two- or three-person Technical Assistance Team (TAT). The team's mandate is to provide States with detailed, readily-implemented recommendations that will lead directly to improved crash data quality measurement and management.

In October 2010, the new contract kicked off with a call to modify the CDIP process, update the pre-site visit materials, and revise the training content. At the request of FHWA and several States, the CDIP team has increased the specificity of the presentations and recommendations –using the State's own data and analyses by preference. Wherever possible, the CDIP recommendations take the form of a worked-out example of how the State should calculate the performance measures recommended by the TAT. If the State has provided the TAT with analytic support or a crash data file that supports calculation of the performance measures, the TAT produces the example performance measures using the State's own data. If the TAT cannot use State data, examples are drawn from other States, including those used in the CDIP presentations. All examples are based on real data.

Beginning in Federal Fiscal Year (FFY) 2013, the National Highway Traffic Safety Administration (NHTSA) joined FHWA on the CDIP management and oversight team. This is a prelude to NHTSA taking full control of the CDIP program in FFY 2014. This change is in parallel with changes NHTSA is implementing in the Traffic Records Advisory and Assessment process. One goal expressed by NHTSA is to use CDIP as one of several types of GO Team support supplied to States at the States' request. In preparation for this transition, the CDIP materials were revised in 2013 to bring them into agreement with the NHTSA Advisory materials so that the same terminology is used in both. In addition, a revised information request and questionnaire now identifies tie-ins to the questions States already answered in the Assessment. In the future, as more States have gone through the new Assessment process, the CDIP team will be able to pre-fill much of the CDIP questionnaire thus reducing the burden placed on States and avoiding having to ask a State for information they have already supplied to NHTSA. States will still need to review and update their answers to the questions.