Timeline for How can I combine fully correlated effect sizes and their standard errors?
Current License: CC BY-SA 4.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| S Apr 26, 2020 at 3:06 | history | bounty ended | CommunityBot | ||
| S Apr 26, 2020 at 3:06 | history | notice removed | CommunityBot | ||
| Apr 24, 2020 at 3:00 | history | tweeted | twitter.com/StackStats/status/1253519144923860993 | ||
| Apr 19, 2020 at 16:09 | answer | added | Haotian Chen | timeline score: 1 | |
| Apr 19, 2020 at 15:26 | comment | added | Speldosa | @carlo It's meant to be indicative of the fact that no measure should be weighted more strongly than any other when it comes to the computation of the final measure. | |
| Apr 19, 2020 at 14:10 | comment | added | carlo | there is no rationale because if you assume something like full correlation, anything can be said about the "real" effect size, it depends on how you justify the difference between measured effect sizes, and it's totally subjective. you could state that the combined ES is the smaller of the two reported, or that it's the bigger one. where does this "full correlation" assumption come from? | |
| S Apr 18, 2020 at 1:57 | history | bounty started | Speldosa | ||
| S Apr 18, 2020 at 1:57 | history | notice added | Speldosa | Draw attention | |
| Apr 15, 2020 at 12:12 | history | asked | Speldosa | CC BY-SA 4.0 |