So on the 28th of September I gave a couple of invited talks at the 2023 JRC Week on Composite Indicators and Scoreboards, whic is spread over a few days and delves into all the technical and non-technical intricacies of building and analysing composite indicators.
I have helped with this training course for some years, particularly when I worked full time in the JRC’s Competence Centre for Composite Indicators and Scoreboards. In this case I gave two lectures:
- Quality control and robustness, in which I explain the main concepts for testing robustness (sensitivity analysis) and give some tips for building high-quality composite indicators.
- A lecture on building and auditing composite indicators with R
For the first lecture you can find the slides here, and the video you can find here (the lecture starts about 14:35).
In the second lecture I built a code demo which is available as an HTML page on GitHub here. In this notebook, I show how to:
- Download data from the Eurostat API in R
- Pass the data and structure into COINr
- Analyse the data and calculate the results
- Map the data interactively using Leaflet
- Explore some methodological variations
- Export everything to Excel
This encapsulates many of the points that I think go in to making a good composite indicator. If you want to hear more about it, check out the recording and skip to approximately 16:28. Unfortunately there were a few technical problems throughout the lecture so you may need to skip ahead on occasions!