Examples

Cytel's LogXact 5 vs. SAS's PROC LOGISTIC:
Relationship between survival and status on the Titanic


 Source: Robert Dawson, Dept.of Mathematics and Computing Science, Saint Mary's University, Halifax, Nova Scotia B3H 3C3, Canada

The records of the sinking of the Titanic were studied to establish the relationship between survival and status on the ship. For each person on board the ocean liner, this dataset records Sex, Age (child/adult), Class (Crew, 1st, 2nd, 3rd Class) and whether or not the person survived.
(2201 observations, 3 covariates, 1 covariate factored.) 


  VARIABLE NAME    DESCRIPTION
  -------------    ---------------------------------------------------------
  CLASS		   Social Class (0 = crew, 1 = first, 2 = second, 3 = third)
  AGE		   Age Group (1 = adult, 0 = child)
  SEX              Gender (1 = male, 0 = female)
  SURV             Survived (1 = yes, 0 = no)
Also, CLASS was specified as a 3-level factor variable.

The results are presented below. Note that SAS fails to fit the logistic regression model -- only LogXact 5 can do it.

LOGXACT 5 RESULTS

=============================================================================================
Parameter Estimates
=============================================================================================
                   Point Estimate            Confidence interval and P-value for Beta
         Type  Beta        SE(Beta)    Type        95.0%     C.I.            Pvalue      SE      
                                                   Lower       Upper       2*1-sided   
=============================================================================================
age       MLE  -1.0615     0.2440      Asymptotic  -1.5398     -0.5833     0.0000              
          CMLE -1.0628     0.2421      Monte Carlo -1.5637     -0.5519     0.0000      0.0006    
                                       ( Seed=1035229319,Samples=10000 )

sex       MLE  -2.4201     0.1404      Asymptotic  -2.6953     -2.1449     0.0000              
          CMLE -2.4104     0.1422      Monte Carlo -2.7014     -2.1338     0.0000      0.0006    
                                       ( Seed=1035229331,Samples=10000 )

class_1   MLE  0.8577      0.1573      Asymptotic  0.5493      1.1661      0.0000              
          CMLE 0.8573      0.1571      Monte Carlo 0.5350      1.1806      0.0000      0.0006    
                                       ( Seed=1035229343,Samples=10000 )

class_2   MLE  -0.1604     0.1738      Asymptotic  -0.5010     0.1802      0.3560              
          CMLE -0.1595     0.1714      Monte Carlo -0.5218     0.1949      0.4016      0.0080    
                                       ( Seed=1035229383,Samples=10000 )

class_3   MLE  -0.9201     0.1486      Asymptotic  -1.2113     -0.6289     0.0000              
          CMLE -0.9174     0.1485      Monte Carlo -1.2217     -0.6221     0.0000      0.0006    
                                       ( Seed=1035229444,Samples=10000 )

%CONST    MLE  2.2477      0.2988      Asymptotic  1.6620      2.8334      0.0000              
=============================================================================================

Analysis time: 00:03:34
=============================================================================================

SAS's PROC LOGISTIC RESULTS

WARNING: There is not enough memory available for exact computations.
Try it yourself!
The links below include data files and SAS code that you can download.

LOGXACT 5--
Download LogXact data
View Instructions for LogXact-5 analysis

PROC LOGISTIC--
Download SAS code (includes data)