Answer:
a) The confidence interval is [tex]0.00\leq\pi_1-\pi_2\leq0.02[/tex].
Step-by-step explanation:
We have to calculate a confidence interval (CI) of a difference of proportions.
For the existing procedure, the proportion is:
[tex]p_1=75/1500=0.05[/tex]
For the new procedure, the proportion is:
[tex]p_2=80/2000=0.04[/tex]
To calculate the CI, we need to estimate the standard deviation
[tex]s=\sqrt{\frac{p_1(1-p_1)}{n_1} +\frac{p_2(1-p_2)}{n_2} }\\\\s=\sqrt{\frac{0.05(1-0.05)}{1500} +\frac{0.04(1-0.04)}{2000} }=\sqrt{ 0.000032 + 0.000019 }=\sqrt{ 0.000051 }\\\\s=0.007[/tex]
For a 90% CI, the z-value is 1.64.
Then, the CI is:
[tex](p_1-p_2)-z*\sigma \leq\pi_1-\pi_2\leq(p_1-p_2)+z*\sigma \\\\(0.05-0.04)-1.64*0.007\leq\pi_1-\pi_2\leq(0.05-0.04)+1.64*0.007\\\\0.01-0.01\leq\pi_1-\pi_2\leq0.01+0.01\\\\0.00\leq\pi_1-\pi_2\leq0.02[/tex]