Data Sources	Coverage Issues	Time Covered	Analysis Objective	Variables	Statistical Technique & Tests	Additional Comments and Assumptions
Kahane	FARS: 1989-1993 fatality data R.L. Polk vehicle registration data (National Vehicle Population Profiles): 1989-1993 11 state accident files, 1989-1993, induced exposure accident involvementsa for data on many control variables	- includes cars & light trucks - includes vehicle occupants and nonoccupants (e.g., pedestrians, bicyclists) - covers all crash modes: rollovers, collisions with objects, pedestrians, big trucks, passenger cars, and light trucks - covers crash avoidance (accident risk per unit of exposure) and crashworthiness (fatality risk given an accident)	Crash experience of MY1985-1993 passenger cars and light trucks during CY1989-1993	To estimate the relationship between vehicle curb weight and fatality risk, per million vehicle exposure years, for MY1985-1993 passenger cars and light trucks based on their crash experience in the U.S. from 1989 through 1993	Dependent Variable: fatality rate per million vehicle years Independent Variables: vehicle curb weight, driver age, sex, vehicle age, state, urban-rural, daytime-nighttime, and other vehicle control variables such as ABS and airbags	Semilog linear aggregate regression (the equation estimates the log-odds of a fatality as a linear function of the independent variables) t test	- The estimates examine the relationship between a vehicle's mass and its overall fatality risk per unit of exposure. The results show the net combined effect of all mass-related crashworthiness and crash avoidance factors; they do not single out whether mass is primarily affecting crashworthiness or crash avoidance (Kahane, p. 13). -The analysis assumes that the distribution of driver age and sex in induced- exposure crashes in the 11-state data base is representative of the general driving public in the U.S. (Kahane, p. 141).
Partykab	NASS 1981-1993 towaway crashes involving moderate to serious driver injuries (i.e., AIS > 2 or fatality)	- includes cars & light trucks - only includes vehicle occupants - only covers vehicle crashworthiness - covers all crash modes except crashes with nonoccupants and principal rollovers where the primary effect of vehicle weight is on crash proneness (i.e., crash avoidance) not crashworthiness	Crash experience of passenger cars and light trucks of all model years during CY1981-1993c	To estimate the relationship between vehicle curb weight and driver risk of moderate to serious injury for all model year passenger cars and light trucks based on their crash experience in the U.S. from CY 1981 through 1993	Dependent Variable: driver injury rate (i.e., no. of injured drivers by AIS level divided by all drivers involved in towaway nonrollover crashes) Independent Variable: vehicle curb weight; vehicles were grouped in 100 lb. increments for cars and 500 lb. increments for light trucks	linear regression (the equation estimates driver injury rates as linear function of vehicle curb weight) t test	Applies driver injury coefficients to estimate total vehicle occupant injuries in 1993
Hertz	State accident data from Illinois (1990-1992) and Florida (1991-1993)	- includes cars and light trucks - only includes vehicle occupants - covers vehicle crashworthiness only - covers all crash modes except principal rollovers and crashes with nonoccupants	crash experience of MY1985-1993 passenger cars and light trucks during CY1990-1992 (Illinois) and CY1991-1993 (Florida)	To estimate the relationship between vehicle weight on per crash incapacitating injury rates for MY1985-1993 passenger cars and light trucks based on their crash experience in two states in 1990 through 1993	Dependent Variable: occurrence or non-occurrence of incapacitating injury Independent Variables: vehicle curb weight, driver age, and geographic location (i.e., rural-urban)--a surrogate for travel speed at the time of the crash	logistic regression (the equation estimates the log-odds of a positive response (i.e., an incapacitating injury) chi square	Estimated change in injury probability for the driver is assumed to be the same for all vehicle occupants

^aVehicles that were standing still (waiting for traffic to clear or a green light) and got hit by somebody else. These involvements are said to be a surrogate for exposure, because they measure how often a vehicle "is there" where it can be hit by other vehicles.

^bThe primary analysis comparable to the Kahane fatality analysis is Partyka, Effect of Vehicle Weight on Crash-Level Driver Injury Rate, June 30, 1995.

^cThere was no NASS estimation file for 1987. The crash analysis time frame was extended, relative to the Kahane fatality analysis, to provide an adequate sample size.