State Crash Data

State Crash Data

A large sample of crash-involved cars is needed to maximize the possibility of detecting any effect of rear-window defoggers. Defoggers may reduce crashes where the driver is backing up and hits something with the rear of the car or when the driver is changing lanes and the car gets hit in the rear. A clear window will obviously help a driver who is backing up, and can also help a driver see if it is safe to change lanes. The goal is to detect if there are proportionately fewer backing up and changing lane crashes involving cars with rear-window defoggers than cars without rear-window defoggers. Since about 50 percent of the model year 1980 domestic cars had rear-window defoggers, the data needed to include a large sample of cars without rear-window defoggers – i.e., 1975-1985 model year vehicles. The data also needed to include specific make, model and model year of vehicles so that the percentage of factory-installed rear window defogger information available in Ward’s Automotive Yearbooks can be linked to the crash data. Therefore, the analyses require a large sample of crash data that includes vehicle maneuver (changing lanes or backing up vs. other), impact location (rear damage), and a large range of model years and VINs or other codes to identify the specific make, model, and model year of cars. State data files are the only data source available to NHTSA that can furnish an adequate number of cases for statistical analysis.

Although the State data files are the best source of data available for this analysis, the State data files may not include all backing up crashes. Backing-up crashes that cause little or no property damage may not be included in a State’s crash file either because it was not reported to the police or it did not meet the State’s minimum-reporting threshold. (As we shall see later on, Pennsylvania has very few backing-up crashes.)

The agency currently receives data from 17 States and maintains these data files for calendar years 1989 and onward for analysis. However, data from 13 States were not used in the analysis: North Carolina, because NHTSA does not have their files prior to 1992; California, Georgia, Illinois, Kansas, New Mexico, Ohio and Washington, because they do not specify the impact location (rear impact); Virginia, because it does not have vehicle make and model; Texas, because it does not have “changing lanes” as a possible vehicle maneuver; Missouri, Utah and Indiana, because earlier years of crash data were not easily available. Florida, Maryland, Michigan and Pennsylvania files were initially considered for the analysis. (As we shall see later on, only the Florida and Michigan data turned out to be suitable for the analysis.)

Previous calendar years of crash data for these four States were obtained to supplement the years currently available at NHTSA. Design Research Engineering supplied 1986–1988 Florida crash data files, the 1986-1988 Maryland crash data files, and the 1981-1988 Michigan crash data. The Pennsylvania Department of Transportation supplied the 1980-1988 Pennsylvania crash data files. Table 7 shows the calendar years of crash data used in the analysis by State. Michigan’s 1992-2000 data files could not be used in the analysis because the make, model and model year were not available. Florida’s 2000 data file was not used in the analysis because it does not contain low-severity non-injury crashes. Pennsylvania’s 1980 file was not used because it would have contributed fewer than 50 cases to the main analysis.

Table 7

Calendar Years Of Crash Data By State

State	Calendar years
Florida	1986-1999
Maryland	1986-2000
Michigan	1981-1991
Pennsylvania	1981-2000