Question 1192817
<font color=black size=3>
At 90% confidence, the z critical value is roughly z = 1.645
Use a Z table to determine this.
Notice that P(-1.645 < Z < 1.645) = 0.90 approximately.


The prior study says phat = 0.67 is the proportion of people who have high speed internet access (i.e. the 67%)
E = 0.08 due to the desired 8% error. The researcher wants to achieve this error or smaller.


The min sample size needed is...
n = phat*(1-phat)*(z/E)^2
n = 0.67*(1-0.67)*(1.645/0.08)^2
n = 93.4847  approximately
n = <font color=red>94</font>  always round UP to the nearest whole number


Let's say we didn't round up and instead went with n = 93 (it's tempting because 93.4847 is closer to 93 than it is to 94)
The margin of error would be roughly
E = z*sqrt(phat*(1-phat)/n)
E = 1.645*sqrt(0.67*(1-0.67)/93)
E = 0.0802
We get an error larger than 8% which isn't what your teacher wants.


However, if we go with n = <font color=red>94</font>, then,
E = z*sqrt(phat*(1-phat)/n)
E = 1.645*sqrt(0.67*(1-0.67)/<font color=red>94</font>)
E = 0.0798
Now the error is under 8%
Likely we won't land *exactly* on 8%, so the next best thing is to get under that value.
The larger n gets, the smaller E will get, and vice versa.


This is why we always round up for minimum sample size problems.


Answer: <font color=red>94</font>
</font>