Selection Criteria for Roll Call Votes

On grounds of inclusion of undesirable votes (type I errors) and exclusion of desirable votes (type II errors), we question the convention of selecting only finalpassage votes for roll call analysis. We propose an alternative selection method based on the estimated salience and strategic significance of roll calls and argue that this method reduces type I and type II errors. We demonstrate that selection of roll calls based on alternative criteria has a major bearing on the asymmetry of partisan roll rates and, we conjecture that its application will also be substantively significant in other modes of inquiry.