Benchmarking using Tidyverse Data Management & Medicare Public Use Files

By Rachel Hayes

We will:

• Download Medicare public use files for part B providers, 2016.
• Do some tidyverse data management.
• Make some scores for the providers – probably using regression.
  Benchmarking.
• How to stratify easily using group_by, nest, and mutate. More than one way to do it.
• How to reshape data using gather and spread.
• How we calculate marginal effect scores to communicate results to end users with no research or statistical background. It’s not ideal, but without it, they lack confidence in the results.