Conceptual
Login

Why a Genetic Correlation Is Not Evidence of a Causal Path

The highest-risk number in the paper for misreading. Educational attainment correlates genetically with pitch variability at r_g = 0.30, and smoking with pitch at r_g = -0.20. A shared genetic signal is consistent with one trait causing the other, with a third trait causing both, and with markers acting on different things for population-structure or social reasons. The node's job is to enumerate those alternatives so the number cannot be retold as 'genes for a lively voice cause more schooling'.

Questions this Concept answers

  • Why does a genetic correlation between two traits fail to establish that one causes the other?