According to an educational film devoted to significance
of tests ("Against All Odds, Inside Statistics" part 20, 1989,
ISBN 1-55946-101-2) the effect of seatbelts on mortality was
actually studied in North Carolina. They refer to 4.7% of deaths
for accidents without seatbelts and 3.8% with seatbels. No other
information was given in the sentence. I assume that the sample
was large enough to make the 0.5% difference significant.
In this case 0.5% --> about 10% reduction in the number of
deaths. Ten percents of ~ 40,000 (deaths per year in the US) is
~ 4000 lives saved per year. I suppose that airbags are much
more effective. Is there an objective evidence for this?
Ludwik Kowalski