The third limitation of election
forensics is that current methods require some fairly restrictive assumptions. These
assumptions pertain to the assumed distribution of vote counts, turnout, and other
indicators in the absence of fraud. This dissertation improves on the tools available to
address this weakness.