In: Statistics and Probability
A television sports commentator wants to estimate the proportion of citizens who "follow professional football." Complete parts (a) through (c). (a) What sample size should be obtained if he wants to be within 2 percentage points with 94% confidence if he uses an estimate of 54% obtained from a poll? The sample size is nothing. (Round up to the nearest integer.) (b) What sample size should be obtained if he wants to be within 2 percentage points with 94% confidence if he does not use any prior estimates? The sample size is nothing. (Round up to the nearest integer.) (c) Why are the results from parts (a) and (b) so close? A. The results are close because 0.54 left parenthesis 1 minus 0.54 right parenthesis equals0.2484 is very close to 0.25. B. The results are close because the confidence 94% is close to 100%. C. The results are close because the margin of error 2% is less than 5%.
z score for 94% confidence interval using the z distribution table is z = 1.88
(A) we have proportion p(hat)= 0.54 and margin of error = 2% = 2/100 = 0.02
Using the sample size formula
setting the given values, we get
this gives us
or
Rounding to next whole number, we get sample size n = 2195
(B) we have margin of error = 2% = 2/100 = 0.02
Sample size formula when no prior estimate for proportion is given
setting the given values, we get
this gives us
or n = 2209
we get sample size n = 2209
(C) We can see from the calculation that 0.54*(1-0.54) gives us 0.2484, which is very close to 0.25(used for no prior estimate)
Thats the only valid reason for sample sizes to be close.
Option B is incorrect because we have used same confidence interval in both part A and B.
option C is incorrect because we have used same margin of error 2% in both part A and B
Only correct answer is option A