Rdperm-Stata Changelog

V.4.0.0 - Oct 2017

Update: added a second rule of thumb to choose q, the effective number of observations from either side of the cutoff. This follows footnote 4 in Canay and Kamat (2017).

Update: updated rule of thumb to choose q, the effective number of observations from either side of the cutoff. We now use the formula in the version of the paper dated May 2017.

New feature: rule of thumb to choose q, the effective number of observations from either side of the cutoff. This choice can be implemented by specifying the option `q(0)'. This is the default option.
New feature: new max test statistic in the construction of the joint test for multiple covariates. This is the default option for the multivariate case. The CvM test statistic can be implemented by specifying the option `cvm'. Note that for a single covariate, these two test statistics are identical.