Skip to main content

Table 4 Percentages of significant cases detected (Detected) and false discovery rate (FDR) after multitest adjustment when the p-values come from families of one-sample t tests where some (% effect) of the alternative hypotheses were true.

From: A new multitest correction (SGoF) that increases its statistical power when increasing the number of tests

 

Strong

  

SB

BH

SGoF

N

% effect

S

Significant

Detected

FDR

Detected

FDR

Detected

FDR

5

5%

100

6.6

0.09

62

0.11

62

0.23

59

5

5%

1000

6.7

0.01

53

0.01

54

0.47

60

5

5%

10000

6.7

0.00

42

0.00

47

1.24

60

5

10%

100

8.5

0.14

31

0.17

30

0.85

46

5

10%

1000

8.4

0.01

36

0.02

35

1.97

43

5

10%

10000

8.3

0.00

31

0.00

30

2.88

45

5

20%

100

11.7

0.17

18

0.28

18

3.03

26

5

20%

1000

11.7

0.02

17

0.03

18

5.28

28

5

20%

10000

11.7

0.00

22

0.00

21

6.23

29

   

Mean

0.05

34.7

0.07

35

2.46

44

   

SD

0.067

15.48

0.099

16.04

2.118

13.91

10

5%

100

8.7

0.49

8

0.75

11

0.95

24

10

5%

1000

8.6

0.09

6

0.27

8

2.25

23

10

5%

10000

8.6

0.01

4

0.11

6

3.21

28

10

10%

100

12.3

1.00

5

2.08

8

3.56

14

10

10%

1000

12.3

0.18

2

1.18

6

5.92

17

10

10%

10000

12.3

0.03

2

0.93

5

6.86

20

10

20%

100

19.5

1.91

3

6.37

5

10.53

9

10

20%

1000

19.6

0.33

1

5.38

4

13.23

11

10

20%

10000

19.6

0.05

1

5.30

4

14.17

12

   

Mean

0.45

3.6

2.49

6.3

6.74

17.6

   

SD

0.631

2.40

2.481

2.29

4.856

6.58

20

5%

100

9.8

2.92

2

4.18

7

1.59

6

20

5%

1000

9.7

1.38

0

4.04

5

3.35

4

20

5%

10000

9.7

0.47

0

3.97

5

4.24

6

20

10%

100

14.4

5.86

1

9.06

5

5.49

2

20

10%

1000

14.3

2.75

0

8.90

5

7.93

3

20

10%

10000

14.3

0.95

0

8.85

4

8.90

5

20

20%

100

23.6

11.78

0

19.04

4

14.64

2

20

20%

1000

23.6

5.50

0

18.88

4

17.23

2

20

20%

10000

23.7

1.91

0

19.01

4

18.27

3

   

Mean

3.72

0.3

10.66

4.7

9.07

3.7

   

SD

3.558

0.71

6.586

0.97

6.209

1.66

  1. Strong: The alternative hypothesis is strong implying that data come from a Normal(0.97, 1). N: sample size. % effect: % of real true alternatives. S: number of tests. Significant: % of significants before adjustment. Detected: % of significant tests after adjustment. FDR: false discovery rate. SB: Sequential Bonferroni. BH: Benjamini and Hochberg. SGoF: Sequential Goodness of Fit.