README.md

This is a repo for an English version of https://github.com/luebby/WWWEKI

Please report any issues here.

Videos: The accompanying expert interviews are availabe from https://wwweki.gitlab.io/interviews/.

Link to course on AI Campus: https://ki-campus.org/courses/whwici

References:

• Rohrer, J. M. (2018). Thinking clearly about correlations and causation: Graphical causal models for observational data. Advances in Methods and Practices in Psychological Science, 1(1), 27-42. https://doi.org/10.1177%2F2515245917745629

• Lübke, K., Gehrke, M., Horst, J., & Szepannek, G. (2020). Why we should teach causal inference: Examples in linear regression with simulated data. Journal of Statistics Education, 28(2), 133-139. https://doi.org/10.1080/10691898.2020.1752859

Tutorials

1. A fork in the road: Walking one way – and not the other (In this module, you will learn: about potential outcomes, counterfactuals, how to define causal effects, and why causal inference is so challenging.)

2. An arrow shows the way (In this module, you will learn: about cause and effect, the basic of causal graphs: the meaning of an arrow, and of parents and children, about causal models, and the difference between observing and doing in the context of causal inference.)

3. Analysing data - with which goal? (In this module you will learn: how to distinguish between description, prediction, and causal inference, why thisd istinction is important and more about the causal ladder.))

4. There is something between us (In this module you will learn: about causal chains, mediators, and that sometimes it is better not to consider certain variables in the analysis.)

5. Storks and babies (In this module you will learn: about causal forks, confounders, and that common causes often lead to confusion.)

6. Kind or handsome? Why not both? (In this module you will learn: about inverted forks, colliders, and that we sometimes unintentionally create associations where there are none.)

7. Why splitting rooms is not a good investment (In this module you will learn: that an observation does not always allow us to derive a suitable action.)

8. Randomness is Magic (In this module you will learn more about: the different data requirements for description and prediction, the advantages of randomly drawn samples, and the advantages of random assignment in the context of experiments.)

9. What would have been, if...? (In this module you will learn: how to determine counterfactuals.)

10. Drawing and reading graphs (In this module you will learn to: draw a graph based on assumptions about the causal structure, use the graph to draw the right consequences for causal inference, run a simulation for the gender pay gap in R.)

11. Does smoking harm adolescents? (In this module you will learn: how to determine a causal effect using linear regression in R based on a real-world example, how to determine which variables need to be adjusted in practice.)

12. Interrogating Data in Practice (In this module you will learn: what critical data interrogation can look like in practice, what else there is to learn about causal inference beyond the basics.)

Licence

Note

This course was supported by a grant from the German Federal Ministry of Education and Research, grant number 16DHBQP040.