#statstab #306 The C-Word: Scientific Euphemisms Do Not Improve Causal Inference From Observational Data

Thoughts: Causal inference is messy business. Maybe we need to be more honest about that.

#causalinference #observational #research #confounds

https://doi.org/10.2105/AJPH.2018.304337

#statstab #273 Statistical Control Requires Causal Justification

Thoughts: Another reason why I just run experiments. Controls are important for causal inference, but what not/to control for is a complex matter.

#causalinference #control #confounds

https://journals.sagepub.com/doi/10.1177/25152459221095823

#statstab #144 Mediation Analysis is Counterintuitively Invalid by @DataColada

Thoughts: Nothing seems less understood or properly used than mediation analysis. Please stop using it.

#mediation #bias #causalinference #confounds

https://datacolada.org/103

#statstab #120 Why do we do matching for causal inference vs regressing on confounders?

Thoughts: A funny and insightful #stackexchange discussion on the pros and cons of causal inference and matching.

#causalinference #stats #matching #confounds

https://stats.stackexchange.com/questions/544926/why-do-we-do-matching-for-causal-inference-vs-regressing-on-confounders/