Dear (younger) colleagues,
Usually, in the textbooks and handbooks it is claimed that the commonly used Pearson correlation is a symmetric measure for correlation because it produces only one estimate regardless of how the variables are declared. However, it is known (although not nesessarily generally) that Pearson correlation has a straight relation to coefficient eta which is a genuinely directional measure. This is generally known by the older generation of researchers, I suppose. However, for the younger generation, I cordially suggest to read a recently published article of this secret characteristics of Pearson correlation and some consequences such as, always, also the explaining powers and effect sizes r2 and R2 are directional measures too.
https://doi.org/10.3389/fpsyg.2022.988660
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression