Timeline for Exceptions to the rule "if the 95% CI does not contain 0, then the effect is significant"
Current License: CC BY-SA 4.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 16, 2023 at 17:31 | comment | added | Sal Mangiafico | @SextusEmpiricus , Another common effect size statistic that is directional but doesn't have 0 as a "no effect" point is odds ratio. But I don't think these are counter examples to the question or my response. For either, we could just change the statement to, "If the CI does not contain 1..." or "contain 0.50". ... The issue is really those statistics (like r-squared), whose CI will never never really cross zero, and --- depending how the confidence interval is calculated --- may practically rarely contain 0, even if there is statistically no effect... | |
| Dec 14, 2023 at 18:30 | comment | added | Sextus Empiricus | "One final note, there's an interesting case of Vargha and Delaney's A, which is directional, but where 0.50 indicates no effect." this counters the entire point that you make in this answer. Yes, often situations where the zero relates to 'no effect' are cases that are directional. But this is not a principle and Vargha and Delaney's A that you mention is a counter example. Other cases are tests for some theoretical prediction different from zero. | |
| Nov 22, 2019 at 15:28 | comment | added | z8080 | THat indeed addresses my question perfectly - thank you! | |
| Nov 22, 2019 at 15:28 | vote | accept | z8080 | ||
| Nov 21, 2019 at 17:15 | history | answered | Sal Mangiafico | CC BY-SA 4.0 |