Ratio Edits Based on Statistical Tolerance Intervals

The role of statistical tolerance intervals for developing ratio edit tolerances in a parametric setup is investigated. The performance of the methodology is assessed for the normal and Weibull distributions. The numerical results show that in terms of Type I and Type II errors, statistical tolerance intervals exhibit better performance compared to other ratio edit procedures available in the literature. The methodology is illustrated using 2010 and 2011 data from the Annual Survey of Manufacturers.