Given suitable data, statistical methods can cover the entire territory, and reach down to every single polling station. However, statistical anomalies can occur for many reasons, not only fraud or error, so positive statistical test results can never be considered a direct proof of irregularities, but only evidence complementing other, qualitative assessments, for instance through observation missions. Manipulation of elections or election results or the occurrence of outcome-altering errors can only be proved through direct evidence, such as a rigorous audit against trustworthy, voter-verified paper records.