Conversely, statistical procedures can miss gross errors and fraud, and are only sensitive to a limited range of problems in the electoral process. Therefore, we cannot conclude from the absence of statistically remarkable anomalies that elections were well administered and free of interference, manipulation, or outcome-altering errors.
