In: Statistics and Probability
An auditor for a hardware store chain wished to compare the efficiency of two different auditing techniques. To do this he selected a sample of nine store accounts and applied auditing techniques A and B to each of the nine accounts selected. The number of errors found in each of techniques A and B is listed in the table below:
Errors in A | Errors in B |
25 | 11 |
28 | 17 |
26 | 19 |
28 | 17 |
32 | 34 |
30 | 25 |
29 | 29 |
20 | 21 |
25 | 30 |
Does the data provide sufficient evidence to conclude that the
number of errors in auditing technique A is greater than the number
of errors in auditing technique B at the 0.1 level of significance?
Select the [Alternative Hypothesis, Value of the Test Statistic].
(Hint: the samples are dependent)
a) [A > B, 1.976]
b) [μD > 0, 1.976]
c) [μD = 0, 1.976]
d) [μD ≥ 0, 1.976]
e) [μD > μ1, 1.976]
f) None of the above
Solution:
An auditor for a hardware store chain wished to compare the efficiency of two different auditing techniques. To do this he selected a sample of nine store accounts and applied auditing techniques A and B to each of the nine accounts selected.
We have to test if the data provide sufficient evidence to conclude that the number of errors in auditing technique A is greater than the number of errors in auditing technique B.
Since samples are dependent, we use paired t test.
d = auditing technique A - auditing technique B
Since we would like to test auditing technique A > auditing technique B,
that is difference is auditing technique A - auditing technique B > 0
thus alternative hypothesis would be:
test statistic value:
where
Thus we need to make following table.
Errors in A | Errors in B | d=A - B | d^2 |
25 | 11 | 14 | 196 |
28 | 17 | 11 | 121 |
26 | 19 | 7 | 49 |
28 | 17 | 11 | 121 |
32 | 34 | -2 | 4 |
30 | 25 | 5 | 25 |
29 | 29 | 0 | 0 |
20 | 21 | -1 | 1 |
25 | 30 | -5 | 25 |
Thus
Thus correct option is:
b) [μD > 0, 1.976]