Correlation is the relatedness of two factors. It's measured by [[Correlation Coefficient]], which ranges from -1 to +1, where 0 is "no correlation". I forget how to do it by hand, but it's one of those things that nobody does by hand. Excel makes it very easy with `correl(...,...)`
# Correlation ≠ Causation
Correlation does not _**necessarily**_ imply causation. Clearly, if A *causes* B you expect to see a correlation, but the simple correlation isn't enough to assume a causal relationship between the variables.
If `A` and `B` are correlated, it could be that:
- `A` causes `B`
- `B` causes `A`
- `A` and `B` are both caused by a 3rd factor
- It's a pure coincidence and they're unrelated[^1]
![[Pasted image 20250118114434.png]]
## Visualizing
Correlations are best seen in [[Scatter Chart]]s, and what [[Pairwise Plot]]s are all about. Also there are [[Fourfold Contingency Table]]s.
Note that [[Statistical Summaries May Hide Truths]].
****
# More
## Source
- [XKCD](https://m.xkcd.com/552/) for the comic
## Related
[^1]: With the advent of [[Big Data]] and the ability to capture and compute millions of factors, you're bound to come across completely unrelated yet very similar looking results.