5 ms·
If I'm reading the scatter plots correctly, it looks like the differences in the means between the two populations is much smaller than the one-sigma variation
by brandmeyer 5y ago
If I'm reading the scatter plots correctly, it looks like the differences in the means between the two populations is much smaller than the one-sigma variation within each population. So while the correlations may technically be statistically significant, they aren't necessarily impactful.
- setgree 5y agoThis is exactly the intuition behind Cohen’s d — the standard unit of comparison for meta-analysis. You take the mean difference between the treatment group and the control group, and divide that number by the standard deviation of the dependent variable, and then assess that result in many studies to see which treatment, or class of treatments, moved the needle the most.
- ChickeNES 5y ago> they aren't necessarily impactful. Much like PLoS ONE ;)