In: Statistics and Probability
A random sample of elementary school children in New York state is to be selected to estimate the proportion p p who have received a medical examination during the past year. It is desired that the sample proportion be within 0.045 of the true proportion with a 98 98 % level of confidence (a) Assuming no prior information about p p is available, approximately how large of a sample size is needed?
If a planning study indicates that pp is around 0.20.2, approximately how large of a sample size is needed?
Solution:
Given that,
a ) = 0.5
1 - = 1 - 0.5 = 0.5
margin of error = E = 0.045
At 98.98% confidence level the z is ,
= 1 - 98.98% = 1 - 0.98.98 = 0.0102
/ 2 = 0.0102/ 2 = 0.0051
Z/2 = Z0.0051 = 2.569
Sample size = n = ((Z / 2) / E)2 * * (1 - )
= (2.569 / 0.045 )2 * 0.5 * 0.5
= 814.80
= 815
n = sample size = 815
b ) = 0.20
1 - = 1 - 0.20 = 0.80
margin of error = E = 0.045
At 98.98% confidence level the z is ,
= 1 - 98.98% = 1 - 0.98.98 = 0.0102
/ 2 = 0.0102/ 2 = 0.0051
Z/2 = Z0.0051 = 2.569
Sample size = n = ((Z / 2) / E)2 * * (1 - )
= (2.569 / 0.045 )2 * 0.20 * 0.80
= 521.47
= 521
n = sample size = 521