StatXact version 6
!Cytel Studio (1.0.0)
>>> DA 

!  Datafile: <new>

!  Table   1 of   1
!------------------------------------------------------!
|    Table1  |     col1    |     col2    |    TOTAL    |
|------------|-------------|-------------|-------------|
|     row1   |          6  |          4  |         10  |
|------------|-------------|-------------|-------------|
|  TOTAL     |          6  |          4  |         10  |
!----------------------------------------!--------------



!One Sample Rates and Proportions:One Binomial
>>> OB BI/EX /SC=3 /CI=0.9200 /MC=0.9900 /CR=10000 /FR=1000 /IM=1000 
/TI=NONE /PR=0.140000 /H1=0.500000 /AL=0.050000 /BStill 


Datafile: <new>

ESTIMATION OF BINOMIAL PARAMETER (PI)

    Number of Trials        =10
    Number of Successes     =4

    Maximum Likelihood Estimate of PI  =      0.4000

    92.00% Confidence Interval for PI:
            (Clopper-Pearson)     = (      0.1400 ,      0.7106)
            (Blyth-Still-Casella) = (      0.1742 ,      0.6891)

    Exact P-values for testing PI        =       0.1400
    One-sided : Pr { T .GE. 4 } =       0.0400
                Pr { T .EQ. 4 } =       0.0326
    Two-sided : 2 * One-sided            =       0.0799


(Note: In this example I have set the confidence coefficient coefficient 
to 92%. At this confidence level the lower confidence bound of the 
Clopper-Pearson confidence interval just touches 0.14. Therefore the 
one-sided p-value for testing the hypothesis that pi=0.14, based on 
Clopper-Pearson, is (1-0.92)/2 = 0.04.)



!One Sample Rates and Proportions:One Binomial
>>> OB BI/EX /SC=3 /CI=0.9600 /MC=0.9900 /CR=10000 /FR=1000 /IM=1000 
/TI=NONE /PR=0.140000 /H1=0.500000 /AL=0.050000 /BStill 


Datafile: <new>

ESTIMATION OF BINOMIAL PARAMETER (PI)

    Number of Trials        =10
    Number of Successes     =4

    Maximum Likelihood Estimate of PI  =      0.4000

    96.00% Confidence Interval for PI:
            (Clopper-Pearson)     = (      0.1138 ,      0.7493)
            (Blyth-Still-Casella) = (      0.1400 ,      0.7106)

    Exact P-values for testing PI        =       0.1400
    One-sided : Pr { T .GE. 4 } =       0.0400
                Pr { T .EQ. 4 } =       0.0326
    Two-sided : 2 * One-sided            =       0.0799


(Note:In this example I have set the confidence coefficient coefficient 
to 96%. At this confidence level the lower confidence bound of the 
Blyth-Still-Casella confidence interval just touches 0.14. Therefore the 
one-sided p-value for testing the hypothesis that pi=0.14, based on 
Blyth-Still-Casella, is (1-0.96)/2 = 0.02.)


Cytel Inc. | 675 Massachusetts Avenue | Cambridge, MA 02139-3309
T: +1.617.661.2011 | F: +1.617.661.4405 | www.cytel.com