The central limit theorem says that, provided an estimator fulfills certain weak conditions, then, for reasonable sample sizes, the sampling distribution of the estimator converges to normality. We propose a procedure to find out what a "reasonably large sample size" is. The procedure is based...

A wildly discussed shortcoming of panel surveys is a potential bias arising from selective attrition. Based on data of the German Socio-Economic Panel Study (SOEP), the authors analyze potential artifacts (level, structure, income inequality) by comparing results for two independently drawn...