Hi!!
We want that the confidence level be 0.99 . So, for a standard normal
variable we need to find Q that:
P(Z > Q) = 0.005 or 0.5%
Due the symmetry of the normal distribution around 0, if Q is such that
P(Z > Q) = 0.005 ,
then we will have that
P(Z < -Q) = 0.005.
Now we can conclude that
P(-Q < Z < Q) = 1 - P(Z > Q) - P(Z < -Q) =
= 1 - 0.005 -0.005 =
= 0.99
This means that the interval (-Q,Q) defines a 99% confidence interval.
Using the table after the following link we find that the value 0,005
correspond to a Z equal to -2.58, this is our -Q, then Q is +2.58:
"Normal Curve Table"
http://www.stat.psu.edu/~rho/stat200/normtable.html
Then for a standard normal variable (-2.58, +2.58) defines a 99%
confidence interval.
Having a large number n of samples from the population (say n > 30)
and knowing Mean and STD (standard deviation) we can calculate the
extremes of the 95% confidence interval by the formula:
Extremes (99% confidence) = Mean +/- (2.58 * STD)/sqrt(n)
For this problem we have that:
n = 90 > 30
Mean = 540 miles
STD = 40 miles
Extreme 1 = 540 + (2.58 * 40)/sqrt(90) = 550.88 miles
Extreme 2 = 540 - (2.58 * 40)/sqrt(90) = 529.12 miles
The 99% confidence interval is (550.88, 529.12) .
I hope that this helps you. Feel free to request for a clarification
if you need it.
Regards.
livioflores-ga |