The inverse cumulative distribution functions return the values at which their respective CDFs attain a given level. This value is typically used in hypothesis testing as a critical value.

There are very few functions for which the inverse CDF can be written in closed form. In most situations the inverse is computed numerically from the CDF.