The Effects of Additive Outliers and Measurement Errors When Testing for Structural Breaks in Variance