Question 1073494
This is a two independent sample test(regular-coffee drinkers vs decaffeinated-coffee drinkers), each sample size is > or = 40, so we can assume a normal distribution
:
We are to test if the mean daily consumption of regular-coffee drinkers(u(1)) is less than that of decaffeinated-coffee drinkers(u(2))
:
This is a one-tailed test, our null and alternate hypothesis tests are
:
Ho: u(2) < or = u(1)
H1: u(2) > u(1)
:
Since we know the standard deviation/s of the population, we can use a z-statistic
:
u(1) of the sample is 4.35 and u(2) of the sample is 5.84
: 
n(1), this sample size is 50 and n(2), this sample size is 40
:
s(1) standard deviation of population is 1.20 and s(2) standard deviation of the population is 1.36
:
z = (5.84 - 4.35) / square root((1.36^2 / 40) + (1.20^2 / 50)) = 5.4393
:
the significance level, alpha, is 0.01
:
z results in a p-value of 0.00001 which is < 0.01, so we reject the null hypothesis(Ho), therefore
:
**********************************************************************
the mean daily consumption of regular-coffee drinkers is greater than that of decaffeinated-coffee drinkers
**********************************************************************