We are always growing our tool set to provide you with anything you might need to design your study. Below is a list of recently added tools and tools that we will be adding soon. If you have any questions, or don't see a tool you need, let us know! Your feed back is important for us to provide you with the support you need.

### Upcoming Functionality:

Export Calculation Reports to PDF
Currently, we offer csv (Excel) downloads for the output of the calculators, randomization list generators and design generators. We are currently working supporting exporting the output to pdf files which will include an overview of the method used and references. In the case of power and sample size it will also include a text snippet describing the result of the calculation for paper and grant submissions.
Currently, our power and sample size calculators assume the population from which the samples are take are infinite. If you know the population however, this has the effect of inflating the sample size/deflating the power of a test. To correct this, you can apply the finite population correction, given by $$\frac{N - n}{N - 1}$$ where $$n$$ and $$N$$ are the sample and total population respectively. We are working on adding the ability to make this correction to all power and sample size calculations.
Early stopping rules and interim analyses are commonly used techniques to ensure that a study is going well or that it is abandoned at the appropriate time if it is going poorly. These rules and analyses need to be planed in advance due to their effect on the overall $$\alpha$$-level of the study. We will be adding support to our calculators to ensure that the study is appropriately planned.

### Upcoming Tools:

Two Sample test of Means with Unequal Variances
Currently, the two sample test of means assumes that the variance of the two samples is the same. This often does not hold up in practice, so we are adding a calculator to deal with the case that the variances are unequal.
Bayesian sample size calculation looks at how many subjects are needed to estimate a given quality of the posterior distribution. One such method finds the sample size needed to find a posterior interval of a given length, $$l$$, under an appropriate coverage criterion and prior distribution. We will be adding calculators that support these methods for one sample and two-sample tests of means.