23 C.F.R. § 1240 app B to Part 1240
A. If State-submitted seat belt use rate information is unavailable or inadequate for both calendar years 1996 and 1997, State seat belt use rates for calendars year 1996 and 1997 will be estimated based on seat belt use rates of fatally-injured occupants. Data from the Fatality Analysis Reporting System (FARS) will be translated into estimated observed seat belt use rates using an algorithm that relates historical belt use by fatally-injured occupants to observed use.
B. The algorithm is as follows:
u = (-.221794 + [RADIC].049193 + .410769F) / .456410
Where:
u = the estimated observed seat belt use
F = the seat belt use in potentially fatal crashes
In the above formula, F is calculated as follows:
F = (f / (1 - e)) / ((f / (1 - e)) + 1 - f)
Where:
F = the seat belt use in potentially fatal crashes
e = State-specific weighted average effectiveness of seat belts in passenger cars and passenger motor vehicles that are not passenger cars
f = State-specific seat belt use rate of fatally-injured occupants of passenger vehicles
C. If State-submitted seat belt use rate information is available for either calendar year 1996 or 1997, but not both, a State seat belt use rate for the year for which information is missing will be estimated by calculating the percent change in the FARS-based observed seat belt use rate (derived from the above algorithm) between the two years. This factor will then be applied to the seat belt use rate from the known year to derive an estimate of the seat belt use rate for the unknown year.
23 C.F.R. § 1240 app B to Part 1240