StatXact 11 versions & prices


incl. 19 % VAT


Data evaluation and quantification of relationships beween data is inevitably based on statistical hypothesis testing and the correct calculation of P-values, which can be laborious, especially in unbalanced or large data sets. By using precise statistics and Monte Carlo simulation, StatXact enables you to solve these problems, even for small data sets and skewed distributions. In addition, the program offers tools to help creating your own exact procedures or combining optimally two or more procedures.

Not only the tools to calculate P-values but also associated methods for sample size calculation (power analysis) are included in the package. The authors of StatXact have not forgotten the user: Although the latest and sometimes complex methods can be used, they are embedded in an easy-to-understand menu system. Thus, StatXact approaches both statisticians in particular and data analysts in general.

Arguments for StatXact:

  • StatXact provides precise statistics
  • StatXact contains the methods of power analysis
  • Easy user interface
  • Exact in respect to small and large samples as well

Software for Small-Sample Categorical and Nonparametric Data

The culmination of ten years of intense research and development, StatXactis the most complete, up-to-date software package for exact nonparametric statistical inference on continuous or categorical data in the world. With StatXact we've covered every mainstream nonparametric procedure, using a full-fledged GUI implementation under Windows. We have provided at least one solution for every important nonparametric statistical problem involving either binary, categorical or continuous data. And, as is our trade-mark, every single one of these solutions is exact. That is, it remains valid for small, sparse and unbalanced data sets. All in all you can access over 80 exact procedures through the StatXact pull-down menus. In addition, we've provided the tools for you to construct your own exact procedures, or to combine two or more procedures in an optimal fashion through a suitable choice of scores. When it comes to handling small-sample data, no other statistical package can even come close to the power and scope of StatXact. Here are some highlights from the StatXact procedures:

  • Exact tests and p-values for Poisson (person-years) data
  • Exact confidence intervals for Poisson (person-years) data
  • Shorter exact confidence intervals for binomial data
  • Exact confidence intervals for multinomial data
  • Conditional (post-hoc) power calculations for stratified linear rank tests
  • Exact confidence intervals for ratios and differences of two binomial proportions
  • Exact Hodges-Lehmann confidence interval for a shift between two distributions
  • Exact goodness-of-fit tests
  • Exact Kolmogorov-Smirnov two-sample test
  • Exact Friedman and Cochran-Q tests for K related samples
  • Exact generalized ANOVA tests
  • Exact tests on measures of association
  • Exact stratified linear rank tests with stratum-specific scores
  • Exact stratified linear rank tests with optimally combined scores
  • Exact K-sample tests for censored survival data

Power Divergence Test
Bayesian Inference
User Defined Tests

Further Information:

StatXact - Trial Version!

You can test the software for free with a 30-days trial version.
You can download the demoversion on the producer website by visiting the link below. You need to fill in a contract. After that the software will be delivered to your E-Mail address as a download.

Download Link

Further Requirements  
Operating System Windows XP, Vista, 7
Min. CPU  
Min. RAM  
Disk Space  

Features of StatXact 11

  • Blaker’s Method to calculate p-vales and confidence intervals
  • Bayesian confidence intervals for the comparison of two proportions with small sample size including risk difference, risk ratio and odds ratio. This includes:
    • Nonparametric -Inference for Categorical data - Two Independent Binomials - Conditional procedures - CI on Odds Ratio
    • Nonparametric - Inference for Categorical data - Two Independent Binomials - Unconditional procedures - CI on difference of proportions
    • Nonparametric - Inference for Categorical data - Two Independent Binomials - Conditional procedures - CI on ratio of proportions
  • User defined tests using R functions
  • New coefficient options for correlated categorical data including two ordered multinomials, unordered RxC tables, singly ordered RxC tables and double ordered RxC tables
  • Parallel Processing for power and sample size calculations
  • Parallel Processing for analysis
  • Correction for discreteness of the binomial distribution in computing confidence intervals

Features of StatXact 10

Windows® 7 and Vista® support
Cytel works continuously to give customers a choice of supported operating systems. Cytel's StatXact® and LogXact® are now both approved for Windows® 7 and Windows® Vista.

R integration

  • Users can now use R scripts in conjunction with Cytel’s exact statistics software. Computers with R 2.3 or newer may run existing files or write and run R programs with the R output displayed in StatXact®.
  • R programs can also run analyses of either StatXact® or LogXact® datasets. All R outputs are displayed in the Cytel Studio interface (except R plots)
  • In addition to R, Cytel continues support of its own batch language to for custom analysis and batch calculations automation.


Tests for two ordered correlated multinomials:
In the Inference for Correlated Categorical Data, Wilcoxon rank sum, Savage scores, Normal scores and Permutation with general scores tests are added for two ordered multinomials.

In the Power and Sample Size module, exact designs for two independent binomials available:

  • Superiority for ratio of proportions (Score test as well as Wald test)
  • Noninferiority for ratio of proportions (Score test as well as Wald test)
  • Equivalence for ratio of proportions (Score test as well as Wald test)
  • The power is computed using exact unconditional type-1 error

Kolmogorov goodness of fit test multinomial distribution:

  • Within a 1xc table data, users can now input either expected count or expected probabilities
  • Range for the parameter Alpha, the value used in computing the confidence interval, has been changed from [0.001,0.999] to [1.e-6,0.999999]

Only StatXact delivers these capabilities

  • R language support to speed analyses and command of batch jobs
  • Stratified Wilcoxon Signed Rank test utilizing case data on paired samples
  • Perform test of correlation across single – not bi-variable – correlation
  • Compute p-value for the test of trend in multiple outcomes
  • Estimate conditional maximum likelihood of trend parameters in correlated data
  • Conduct clustered response tests for correlation significance