A Louis Harris and Associates survey conducted in 1984 for Philip Morris, Inc..
ID: 3175454 • Letter: A
Question
A Louis Harris and Associates survey conducted in 1984 for Philip Morris, Inc.. showed an increase over the decade in attendance at movies, concerts, and the theater. The poll was based on a telephone survey of 1504 adults, and the results were compared with similar polls conducted in 1973. 1975. and 1980 (Wall Street Journal, December 1, 1984). If the survey in 1973 was based on an equal number of interviews, how accurately could you estimate the difference in attendance proportions between 1973 and l984? Suppose that you were designing the surveys and wished to estimate the difference in a pair of proportions. correct to within .02. with probability .9. How many interviewees should be included in each sample?Explanation / Answer
a ) we need to calculate the margin of error here which is given as
Z0.05/2*(p*(1-p)/n)1/2 where p =0.75 , so putting the value of n = 1500 and p = 0.75 and Z as 1.96 , assuming 95% confidence interval we get , we get z = 1.96 from the z table
Z0.025*(0.75*0.25/1500)1/2 => 1.96*(0.1875/1500)1/2 solving this we get = 0.0219 or 2.1%
b ) in the second part using the same equation , we are given the error as 0.2 and p as 0.9 and we need to calculate the value of N
Z0.025*(0.09/n)1/2 = 0.02 => 1.96*(0.09/n)1/2 = 0.02
.