From the problem, we are given
x1 - x 2 = 10 cm
s of (x1 - x2)i = 6 cm
The null hypothesis is that there is no significant difference between the height of males of the two neighborhoods.
For the null hypothesis to be rejected, this condition must be met
u1 - u2 ≠ 0
For a normal distribution
u = x + z s / √n
where z depends on the confidence interval (CI)
If
CI = 67%, z = 0.16
CI = 95%, z = 0.025
CI = 99.7%, z = 0.0015
CI = 100%, z = 0
So,
u1 - u2 = x1 + z s1 √n - (x2 - z s2 /√n )
Substituting the given values and trying out the different values of z for different confidence intervals, the CI that would meet the required condition is
CI = 95%
So, the answer is
The null hypothesis must be rejected if we choose the 95% confidence level.