Question 314394
Depends on what you are trying to estimate, the population average (mean) speed of an individual car or the populatin average (mean) speed for samples of size 36.


If we assume the speeds are normally distributed, then
for estimating the average speed of individual cars the formula would be

(Xbar - Z*S, Xbar + Z*S)  where S=standard deviation
for 95% confidence interval Z=1.96
(60-1.96*6, 60+1.96*6) or  60-11.8=48.2, 60+11.2-71.2  

For estimating the average speed for groups of size 36, in theory you should use the t distribution because your sample is small, and use the Standard Error not the standard deviation.  Standard Error = Standard deviation/Sqrt(n)

with n=36 and 95% confidence 
t at 0.05/2=0.025 and 35 degrees of freddom =2.03 
SE = 6/Sqrt(36) =1

Xbar-t*SE, Xbar + t*SE  which is (60-2.03*1, 60+2.03*1) or (57.97, 62.03)

Some people say that if n>30 you can use Z instead of t, but this is not correct since Z and t do not merge until n>200

if you did use Z, then for 95% confidence Z=1.96
(60-1.96*1, 60+1.96*1) or (58.04, 61.96)