Does load "A" shoot
groups of different size than load "B"?
Enter the group
sizes in "A" and "B", then enter the rank of each group in "Rank A" or
"Rank B". Then add up the ranks and put the total in "Rank Sum".
|
|
A |
Rank A |
B |
Rank B |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Rank
Sum>>>>>>> |
|
|
|
|
(If there is a
"tie", for example, if there are two identical values that would be third
and fourth, then assign each a rank of 3.5.)
How many entries are
there in "A" and "B"? "A"_____, "B"_____
Now examine the
table, the rank sums, and the number of entries in "A" and "B".
|
Upper and lower
bounds, Wilcoxon Rank Sum Test |
|
|
|
|
|
|
|
|
|
|
|
95% sure |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
n |
|
3 |
3 |
|
4 |
4 |
|
5 |
5 |
|
6 |
6 |
|
7 |
7 |
|
8 |
8 |
|
9 |
9 |
|
10 |
10 |
|
|
|
L |
U |
|
L |
U |
|
L |
U |
|
L |
U |
|
L |
U |
|
L |
U |
|
L |
U |
|
L |
U |
|
3 |
|
5 |
16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
|
6 |
18 |
|
11 |
25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
|
6 |
21 |
|
12 |
28 |
|
18 |
37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
|
7 |
23 |
|
12 |
32 |
|
19 |
41 |
|
26 |
52 |
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
|
7 |
26 |
|
13 |
35 |
|
20 |
45 |
|
28 |
56 |
|
37 |
68 |
|
|
|
|
|
|
|
|
|
|
8 |
|
8 |
28 |
|
14 |
38 |
|
21 |
49 |
|
29 |
61 |
|
39 |
73 |
|
49 |
87 |
|
|
|
|
|
|
|
9 |
|
8 |
31 |
|
15 |
41 |
|
22 |
53 |
|
31 |
65 |
|
41 |
78 |
|
51 |
93 |
|
63 |
108 |
|
|
|
|
10 |
|
9 |
33 |
|
16 |
44 |
|
24 |
56 |
|
32 |
70 |
|
43 |
83 |
|
54 |
98 |
|
66 |
114 |
|
79 |
131 |
If
either
rank sum is equal to or between the table numbers, we are
not
95% sure that load "A" produces different accuracy than load "B".
If
neither
rank sum is equal to or between the two table numbers, we
are
95% sure that load "A" and load "B" produce different accuracy.
Reference "Chapter
7.4 ON ACCURACY"
