Using Statistics To Make Inferences 6 1
A novel treatment was compared against control values.
Treated Control
236.9 400.6 403.3 509.3 562.6 664.4 747.7 795.8 962.9 1146.4 1238.9 308.5 321.5 397.8 400.4 466.2 580.0 887.1
Perform a Mann-Whitney Test to assess the efficacy of the treatment. 1
Note that the original data is sorted to aid the calculations.
Data 236.90 308.50 321.50 397.80 400.40 400.60 403.30 466.20 509.30 562.60 580.00 664.40 747.70 795.80 887.10 962.90 1146.40 1238.90
Source Treated Control Control Control Control Treated Treated Control Treated Treated Control Treated Treated Treated Control Treated Treated Treated
Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Treated 1
Total
123
Control 2 3 4 5
6 7 8 9 10 11 12 13 14 15 16 17 18 48
nT nT 1 11 (11 1) RT 11 7 123 20.0 (W = 123) 2 2 n n 1 7 (7 1) nT nC C C RC 11 7 48 57.0 2 2
U T nT nC UC
(mid-point 1 nT nC 38.5 so only need calculate one) 2 For nT 11 , nC 7 , the critical value from the tables for p=0.025 is 16 . The result is not significant at the 5% level, the two samples do not appear to differ. For p=0.05 it is 19 , the result is not significant at the 10% level. Close to the value U (see 0.1031 in the test reported).
z
1 nT nC 2 2 1.63 nT nC nT nC 1 12 U
Which is greater than -1.96, there does not appear to be a significant difference.
Mike Cox
6.1
Version 1
MTB > Mann-Whitney 95.0 'Treated' 'Control'; SUBC> Alternative 0. Mann-Whitney Confidence Interval and Test Treated N = 11 Median = 664.4 Control N = 7 Median = 400.4 Point estimate for ETA1-ETA2 is 198.2 95.4 Percent CI for ETA1-ETA2 is (-70.8,496.7) W = 123.0 Test of ETA1 = ETA2 vs ETA1 not = ETA2 is significant at 0.1031 Cannot reject at alpha = 0.05
2
A random sample of 15 male and 20 female students had their academic ability tested. The scores are displayed below. Men
10.00 15.20 20.00
11.00 16.00
12.00 17.00
12.40 18.50
12.50 19.00
12.50 19.25
14.00 19.50
Women 15.00 18.25 20.00
16.20 18.30 21.00
16.30 18.50 21.10
17.25 18.75 23.00
17.50 19.10 24.00
17.75 19.20 25.00
18.00 19.75
From the results would you conclude that there is no difference between the two groups (The samples are from populations with similar medians)? 2
A Mann-Whitney test is appropriate, since there is no reason to assume the scores are normally distributed. The data is already sorted. The only problem is ties. Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Mike Cox
True rank 1 2 3 4 5.5 5.5 7 8 9 10 11 12 13 14 15 16 17 18 19 20.5 20.5 22 23 24 25 26 27 28 29.5
6.2
Score 10.00 11.00 12.00 12.40 12.50 12.50 14.00 15.00 15.20 16.00 16.20 16.30 17.00 17.25 17.50 17.75 18.00 18.25 18.30 18.50 18.50 18.75 19.00 19.10 19.20 19.25 19.50 19.75 20.00
Men Men Men Men Men Men Men Women Men Men Women Women Men Women Women Women Women Women Women Men Women Women Men Women Women Men Men Women Men
True rank 1 Men 2 Men 3 Men 4 Men 5.5 Men 5.5 Men 7 Men 9 Men 10 Men 13 Men 20.5 Men 23 Men 26 Men 27 Men 29.5 Men 8 Women 11 Women 12 Women 14 Women 15 Women 16 Women 17 Women 18 Women 19 Women 20.5 Women 22 Women 24 Women 25 Women 28 Women
Version 1
Sum 186
30 31 32 33 34 35
29.5 31 32 33 34 35
20.00 21.00 21.10 23.00 24.00 25.00
Women Women Women Women Women Women
29.5 31 32 33 34 35
Women Women Women Women Women Women
Sum 444
Here nM = 15, nW = 20 with SM = 186 and SW = 444 UM = nM nW + ½ nM (nM + 1) - SM = 234 and UW = nM nW + ½ nW (nW + 1) - SW = 66
U = min(UM,UW) = 66 and z
1 nM nW 2 2 2.78 nM nW nM nW 1 U
12 Which is less than -1.96, there appears to be a significant difference. A check employing Minitab, plus a two sample t test, just in case. ROW
Men
Women
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
10.00 11.00 12.00 12.40 12.50 12.50 14.00 15.20 16.00 17.00 18.50 19.00 19.25 19.50 20.00
15.00 16.20 16.30 17.25 17.50 17.75 18.00 18.25 18.30 18.50 18.75 19.10 19.20 19.75 20.00 21.00 21.10 23.00 24.00 25.00
Mann-Whitney Confidence Interval and Test Men N = 15 Median = 15.200 Women N = 20 Median = 18.625 Point estimate for ETA1-ETA2 is -4.000 95.3 Percent C.I. for ETA1-ETA2 is (-6.250,-1.249) W = 186.0 Test of ETA1 = ETA2 vs. ETA1 = ETA2 is significant at 0.0054 The test is significant at 0.0054 (adjusted for ties) Men Women
N 15 20
MEAN 15.257 19.198
MEDIAN 15.200 18.625
TRMEAN 15.296 19.108
Men Women
MIN 10.000 15.000
MAX 20.000 25.000
Q1 12.400 17.562
Q3 19.000 20.750
STDEV 3.442 2.581
TWOSAMPLE T FOR Men VS Women
Mike Cox
6.3
Version 1
SEMEAN 0.889 0.577
Men Women
N 15 20
MEAN 15.26 19.20
STDEV 3.44 2.58
SE MEAN 0.89 0.58
95 PCT CI FOR MU Men - MU Women: ( -6.12,
-1.76)
TTEST MU Men = MU Women (VS NE): T= -3.72 P=0.0010 DF= 25
3
A company run a chain of chemists. A female pharmacist working for the chain sued alleging discrimination in the promotion from pharmacist to chief pharmacist. Given below are the data showing the time from initial employment to chief pharmacist in months. Female 229
453
Male 5 34
3
7 37
12 47
14 49
14 64
14 67
18 69
21 125
22 192
23 483
24
25
Does the evidence support the claim? The details of the Mann-Whitney calculation are shown below. Score 5 7 12 14 14 14 18 21 22 23 24 25 34 34 37 47 49 64 67 69 125 192 229 453 483
Rank Female Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Female Female Male Sum U mean var. z
Mike Cox
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 47 278 2 44 23.00 23.00 99.67 99.67 -2.104 2.10
6.4
Rank Female Male reversed 25 25 24 24 23 23 22 22 21 21 20 20 19 19 18 18 17 17 16 16 15 15 14 14 13 13 12 12 11 11 10 10 9 9 8 8 7 7 6 6 5 5 4 4 3 3 2 2 1 1 5 44 23.00 99.67 2.10
320 2 23.00 99.67 -2.104
Version 1
34
Here nM = 23, nF = 2 then SM = 278 and SF = 47 UM = nM nF + ½ nM (nM + 1) - SM = 44 and UF = nM nF + ½ nF (nF + 1) - SF = 2
1 nM nF 2 2 2.053 n M n F n M n F 1 12 U
z
Note that ties need not be considered in detail since they occur in the same sample. There would appear to be a clear case of discrimination, when compared to the critical value of ±1.96 for a difference.
A Mann-Whitney test is expected, however just in case a two sample t test is attempted some basic statistics are also given.
Data Display Female 229
453
Data Display Male 5 34
7 37
12 47
14 49
14 64
14 67
18 69
21 125
22 192
23 483
24
Median 341 25.0
Q3 * 64.0
25
34
Descriptive Statistics: Female, Male Variable Female Male
N 2 23
N* 0 0
Mean 341 60.9
SE Mean 112 21.1
StDev 158 101.4
Minimum 229 5.0
Q1 * 14.0
Maximum 453 483.0
Two-Sample T-Test and CI: Female, Male Two-sample T for Female vs Male Female Male
N 2 23
Mean 341 61
StDev 158 101
SE Mean 112 21
Difference = mu (Female) - mu (Male) Estimate for difference: 280.130 95% CI for difference: (-1168.076, 1728.337) T-Test of difference = 0 (vs not =): T-Value = 2.46
P-Value = 0.246
DF = 1
Mann-Whitney Test and CI: Female, Male Female Male
N 2 23
Median 341.0 25.0
Point estimate for ETA1-ETA2 is 223.0 96.0 Percent CI for ETA1-ETA2 is (36.9,441.1)
Mike Cox
6.5
Version 1
W = 47.0 Test of ETA1 = ETA2 vs ETA1 not = ETA2 is significant at 0.0400
The test is significant at 0.0398 (adjusted for ties)
Mike Cox
6.6
Version 1
Critical Values For Wilcoxon's Signed-Rank Test The body of the table contains the critical values for Wilcoxon's signed-rank test. Always enter the table with W+, the sum of the ranks of the positive deviations. If a critical value is missing, the hypothesis can not be rejected for this combination of n and α. One Tail Probability 0.05
0.025
0.01
OneTail Probability
0.005
n
0.05
Two Tail Probability
Mike Cox
n
0.10
0.05
0.02
5
1
6
2
1
7
4
2
0
8
6
4
2
9
8
6
10
11
11
0.025
0.01
0.005
TwoTail Probability
0.01
n
0.10
0.05
0.02
0.01
28
130
117
102
92
29
141
127
111
100
30
152
137
120
109
0
31
163
148
130
118
3
2
32
175
159
141
128
8
5
3
33
188
171
151
138
14
11
7
5
34
201
183
162
149
12
17
14
10
7
35
214
195
174
160
13
21
17
13
10
36
228
208
186
171
14
26
21
16
13
37
242
222
198
183
15
30
25
20
16
38
256
235
211
195
16
36
30
24
19
39
271
250
224
208
17
41
35
28
23
40
287
264
238
221
18
47
40
33
28
41
303
279
252
234
19
54
46
38
32
42
319
295
267
248
20
60
52
43
37
43
336
311
281
262
21
68
59
49
43
44
353
327
297
277
22
75
66
56
49
45
371
344
313
292
23
83
73
62
55
46
389
361
329
307
24
92
81
69
61
47
408
379
345
323
25
101
90
77
68
48
427
397
362
339
26
110
98
85
76
49
446
415
380
356
27
120
107
93
84
50
466
434
398
373
6.7
Version 1
Critical Values Of U In The Mann-Whitney Test Critical Values of U at α = 0.025 with direction predicted or at α = 0.05 with direction not predicted.
n2
9
10
11
12
13
14
15
16
17
18
19
20
2
0
0
0
1
1
1
1
1
2
2
2
2
3
2
3
3
4
4
5
5
6
6
7
7
8
4
4
5
6
7
8
9
10
11
11
12
13
13
5
7
8
9
11
12
13
14
15
17
18
19
20
6
10
11
13
14
16
17
19
21
22
24
25
27
7
12
14
16
18
20
22
24
26
28
30
32
34
8
15
17
19
22
24
26
29
31
34
36
38
41
9
17
20
23
26
28
31
34
37
39
42
45
48
10
20
23
26
29
33
36
39
42
45
48
52
55
11
23
26
30
33
37
40
44
47
51
55
58
62
12
26
29
33
37
41
45
49
53
57
61
65
69
13
28
33
37
41
45
50
54
59
63
67
72
76
14
31
36
40
45
50
55
59
64
67
74
78
83
15
34
39
44
49
54
59
64
70
75
80
85
90
16
37
42
47
53
59
64
70
75
81
86
92
98
17
39
45
51
57
63
67
75
81
87
93
99
105
18
42
48
55
61
67
74
80
86
93
99
106
112
19
45
52
58
65
72
78
85
92
99
106
113
119
20
48
55
62
69
76
83
90
98
105
112
119
127
n1
Mike Cox
6.8
Version 1
Critical Values Of U In The Mann-Whitney Test Critical Values of U at α = 0.05 with direction predicted or at α = 0.10 with direction not predicted.
n2
9
10
11
12
13
14
15
16
17
18
19
20
2
1
1
1
2
2
2
3
3
3
4
4
4
3
3
4
5
5
6
7
7
8
9
9
10
11
4
6
7
8
9
10
11
12
14
15
16
17
18
5
9
11
12
13
15
16
18
19
20
22
23
25
6
12
14
16
17
19
21
23
25
26
28
30
32
7
15
17
19
21
24
26
28
30
33
35
37
39
8
18
20
23
26
28
31
33
36
39
41
44
47
9
21
24
27
30
33
36
39
42
45
48
51
54
10
24
27
31
34
37
41
44
48
51
55
58
62
11
27
31
34
38
42
46
50
54
57
61
65
69
12
30
34
38
42
47
51
55
60
64
68
72
77
13
33
37
42
47
51
56
61
65
70
75
80
84
14
36
41
46
51
56
61
66
71
77
82
87
92
15
39
44
50
55
61
66
72
77
83
88
94
100
16
42
48
54
60
65
71
77
83
89
95
101
107
17
45
51
57
64
70
77
83
89
96
102
109
115
18
48
55
61
68
75
82
88
95
102
109
116
123
19
51
58
65
72
80
87
94
101
109
116
123
130
20
54
62
69
77
84
92
100
107
115
123
130
138
n1
Mike Cox
6.9
Version 1