you want to do a more detailed analysis of subgroups
less accuracy is acceptable
a higher level of confidence is required
there is more variability in the population
If you decide to use a 95% confidence interval rather than a 99% confidence interval you would expect the confidence interval to become
wider
stay the same
smaller
increase by 4%
Given the sample statistics n = 80, = £9.90 and s = £1.84 , the 95% confidence interval is given by
μ = £9.90 ± £0.40
μ = £9.90 ± £1.84
μ = £8.06 ± £1.84
μ = £9.90 ± £3.61
To estimate the true mean to within £5.00 with a 95% confidence interval, given a standard deviation of £56.50 requires a sample size of
100
250
500
1000
In a survey of 300 call customers, 33 complain that connection time is a problem for them. The 95% confidence interval is
33 ± 2%
33 ± 4%
11% ± 2%
11% ± 4%
Assuming a ‘worse case scenario’ that p = 50% the sampling size required to give a 95% confidence interval with a sampling error of no more that ± 2% is
1000
1500
2000
2500
Given n1 = 150, x1 = £78.40, s1 = £8.20 and n2 = 180, x2 = £72.00, s2 = £6.45 the 95% confidence interval for the difference of means is given by
£6.40 ± £1.62
£6.40 ± £1.96
£6.40 ± £0.82
£6.40 ± £3.24
Given n1 = 150, p1 = 7% and n2 = 180, p2 = 4% the 95% confidence interval for the difference of means is given by
3% ± 3%
3% ± 5%
3% ± 7%
3% ± 9%
A sample of 70 is taken from a population of 80. The finite population correction factor is
0.9354
0.875
0.3536
none of these
A random sample of 8 business transactions was selected from a large number of such transactions. The degrees of freedom and the critical value using the t-distribution for the 95% confidence interval would be