Answer: a. n= 1068
b. n= 164
Step-by-step explanation:
The formula to find the sample size :
[tex]n=p(1-p)(\dfrac{z^{*}}{E})^2[/tex]
, where p=prior population proportion , z* = critical z-value and E = Margin of error.
Here , let p=proportion of computers that use a new operating system.
Given : Confidence level = 95%
i.e. z* = 1.96 [by z-table]
Margin of error : E = 3% =0.03
a. If p is unknown , then we assume p=0.5
Then, [tex]n=(0.5)(1-0.5)(\dfrac{1.96}{0.03})^2=1067.11111\approx1068[/tex]
i.e. n= 1068
b. p=0.96
Then, [tex]n=(0.96)(1-0.96)(\dfrac{1.96}{0.03})^2=163.908266667\approx164[/tex]
i.e. n= 164.