1 Load packages 2 Motivation 3 Computing the effective sample size in the 2016’ US federal elections 4 Conclusion 5 Further reading 6 Reproducibility 1 Load packages library(tidyverse) # data wrangling 2 Motivation My colleague, Karsten Lübke, first grade statistician, pointed me out to a paper … In 2018, the statistican Meng wrote a paper about biases in big data see here. In a nutshell, he argues that non-random samples will be worse when data is …