Leaders confuse correlation with causation; they assume actions for which they are responsible caused any improvements in performance, including those best explained by regression to the mean and those that are contradicted by other measures.