Robust and accurate empirical evidence on the social world can empower organizations to make effective decisions, gain a better understanding of their markets, and develop successful products. This can be achieved with rigorous data collection, deep understanding of the biases that can emerge in observational data, and careful application of suitable statistical methods to analyze the data. I advice companies on how to collect, correct, measure and analyze their data. This includes:   
  • Accuracy and reliability assessment of large scale-human labeling projects​
  • Bias assessment in ML models
  • Measurement design
  • Survey and sampling design 
  • Statistical analyses of observational data