Data with observations at different points in time lead to limitations in applying conventional statistical modeling and inference that depends on the assumptions that the adjacent observations are independent and identically distributed. Seat belts policies were evaluated by using time series intervention analysis. The introduction of seat belt law in Jordan resulted in reducing casualties by 1 unit during the first six month transition period. This was followed by a reduction of 3 units in the number of casualties during the period of complete enforcement of seat belt usage on the highways and within city limits. An increase of 77 units in the number of casualties was obtained during the obtained during the period of enforcing seat belt usage on roads outside city limits in Jordan. The overall effect of seat belt policies on casualties was reducing the number of casualties by 2 units. Seat belt effects on fatalities and injuries are evaluated. This new method of analysis clearly shows that the second period of seat belt enforcement is the most effective in reducing casualties, injuries and fatalities. (A)
Abstract