b)The research director of a large supermarket chain wished to measure the effect of a storewide sales campaign on sales of non-sale items and believes that sales of non-sale items may also be affected by the overall average weekly sales volume of the store.

He randomly selected 10 stores and then, another 10 stores that matches the first 10 according to average weekly sales volume. The results, after pairing, are shown in the following table

Matched Pair

Sales of non-sale items - with sales campaign, (x)

Sales of non-sale items - without sales campaign, (y)

Differences

d =(x - y)

Line 1 Matched Pair

1

2

3

4

5

6

7

8

9

10

Line 2-Sales of non-sale items - with sales campaign, (x)

67.2

59.4

80.1

38.4

49.3

94.7

31.7

64.3

57.3

47.6

Line 3 -Sales of non-sale items - without sales campaign, (y)

65.3

54.7

81.3

37.9

41.7

89.0

33.0

58.4

52.4

39.8

Line 4-Differences

d =(x - y)

1.9

4.7

-1.2

0.5

7.6

5.7

-1.3

5.9

4.9

7.8

TotalLine2 total590Line3 total553.5

Assuming that all sales of non-sale items are normally distributed regardless of the campaign, use the above information to test whether sales of non-sale items are significantly higher with the sales campaign than without. Take a = 0.05.

Note: You may use the following summary values when answering this question - The sample of differences in the above table has a mean of 3.65 and a standard deviation of 3.428.

I couldnot copy the tables correctly so I have mentioned line number so that its easy for you work on.