The tipr R package has some new features! And a new and improved API! What is tipr tipr is an R package that allows you to conduct sensitivity analyses for unmeasured confounders. Why might you want to do that? Well, as it turns out, the assumption of “no unmeasured confounders” is integral to any estimation of a causal effect. This assumption is untestable, so often the best we can do is examine how far off our estimates would be should an unmeasured confounder exists, hence sensitivity ...
