Skip to main content

Table 5 All possible outcomes, p-values, and corresponding decisions for an RCT with n = 8 patients (4 per treatment arm) with Fisher’s exact test

From: A roadmap to using randomization in clinical trials

Number of responders

Difference in proportions (Experimental vs. Control)

Fisher’s exact test 1-sided p-value

Decisiona

Experimental

Control

0/4

0/4

0

1.0

F

1/4

1/4

0

0.7857

F

2/4

2/4

0

0.7571

F

3/4

3/4

0

0.7857

F

4/4

4/4

0

1.0

F

1/4

0/4

0.25

0.5

F

2/4

0/4

0.50

0.2143

F

3/4

0/4

0.75

0.0714

S

4/4

0/4

1

0.0143

S

0/4

1/4

-0.25

1.0

F

0/4

2/4

-0.50

1.0

F

0/4

3/4

-0.75

1.0

F

0/4

4/4

-1

1.0

F

2/4

1/4

0.25

0.5

F

3/4

1/4

0.50

0.2429

F

4/4

1/4

0.75

0.0714

S

1/4

2/4

-0.25

0.9286

F

1/4

3/4

-0.50

0.9857

F

1/4

4/4

-0.75

1.0

F

3/4

2/4

0.25

0.5

F

4/4

2/4

0.50

0.2143

F

2/4

3/4

-0.25

0.9286

F

2/4

4/4

-0.50

1.0

F

4/4

3/4

0.25

0.5

F

3/4

4/4

-0.25

1.0

F

  1. a F Declare study a failure, S Declare study a success