March 2026
Beginner
212 pages
5h 54m
English
After all the things we have studied together, let me take you through a last twist: causal estimates can be misleading. How is this possible? The main reason is that we tend to overinterpret the results or have blind spots due to the choice of metric used for the outcome. First, measuring a causal effect does not automatically explain how the effect happened. In a randomized experiment to test the effect of cold showers on health, you might be able to measure a causal effect on work-related sick leave. But is it due to an immune boost from temperature shock? Is it a mental resilience effect that lowers stress? Is it just a placebo effect? As we have seen previously, ...
Read now
Unlock full access