In: Statistics and Probability
A simple random sample of front-seat occupants involved in car crashes is obtained. Among 2771occupants not wearing seat belts, 38 were killed. Among 7702
occupants wearing seat belts, 15 were killed. Use a 0.01 significance level to test the claim that seat belts are effective in reducing fatalities. Complete parts (a) through (c) below.
a) identify test statistic
b) identify the p-value
c) identify confidence interval
= 38/2771 = 0.0137
= 15/7702 = 0.0019
a) The pooled sample proportion(P) = ( * n1 + * n2)/(n1 + n2)
= (0.0137 * 2771 + 0.0019 * 7702)/(2771 + 7702)
= 0.005
The test statistic z = ()/sqrt(P(1 - P)(1/n1 + 1/n2))
= (0.0137 - 0.0019)/sqrt(0.005 * (1 - 0.005) * (1/2771 + 1/7702))
= 7.55
b) P-value = P(Z > 7.55)
= 1 - P(Z < 7.55)
= 1 - 1 = 0
c) At 99% confidence interval the critical value is z0.005 = 2.58
The 99% confidence interval is
() +/- z0.005 * sqrt(P(1 - P)(1/n1 + 1/n2))
= (0.0137 - 0.0019) +/- 2.58 * sqrt(0.005 * (1 - 0.005) * (1/2771 + 1/7702))
= 0.0118 +/- 0.004
= 0.0078, 0.0158