Timeline for Correlation confidence interval: difference between standard calculation and bootstraping
Current License: CC BY-SA 4.0
17 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 15, 2020 at 16:15 | answer | added | EdM | timeline score: 4 | |
| Jul 15, 2020 at 13:44 | answer | added | user289381 | timeline score: 3 | |
| Jul 15, 2020 at 12:31 | history | reopened |
StupidWolf Dan Chaltiel whuber♦ |
||
| Jul 15, 2020 at 10:04 | comment | added | StupidWolf | here's also an article written by a user on CV :) stata-journal.com/sjpdf.html?articlenum=pr0041. If I am not wrong, it is the sample size. If you have a larger dataset, it should be pretty similar. | |
| Jul 15, 2020 at 10:04 | comment | added | StupidWolf | if you check the source code for cor.test, it does a fisher transformation on the correlation coefficient, and from there estimate a confidence interval and back transform. Hence you see the asymmetric distribution. stats.stackexchange.com/questions/272434/… | |
| Jul 15, 2020 at 9:42 | comment | added | Dan Chaltiel | @StupidWolf thanks for the suggestion, I added the table | |
| Jul 15, 2020 at 9:41 | history | edited | Dan Chaltiel | CC BY-SA 4.0 |
added 475 characters in body
|
| Jul 15, 2020 at 9:32 | comment | added | StupidWolf | hi @DanChaltiel i think it's ok after the explanation, it explains the plots better especially for people who might know R but not familiar with tidyverse. One suggestion I can also offer is to avoid this back and forth, is to put a table of your estimates and upper / lower bound of CI estimates | |
| Jul 15, 2020 at 7:17 | review | Reopen votes | |||
| Jul 15, 2020 at 12:31 | |||||
| Jul 15, 2020 at 6:58 | comment | added | Dan Chaltiel | @whuber is that OK for you after my edit? | |
| Jul 15, 2020 at 6:57 | history | edited | Dan Chaltiel | CC BY-SA 4.0 |
added 868 characters in body
|
| Jul 14, 2020 at 16:14 | comment | added | whuber♦ |
I agree that it doesn't seem to be a programming question. I am primarily concerned (therefore) with ensuring that the question is understandable even by people not conversant with R and the R packages you are employing.
|
|
| Jul 14, 2020 at 13:55 | comment | added | Dan Chaltiel |
@whuber I added an r tag, as this question may be a bit dependant on the statistical behavior of R itself (the tag description fits perfectly to the question). I will try to explain my code with comments but I'm afraid that knowing R might be somehow mandatory to answer this question. Still, this is not a programming question at all and I really think that it belongs here on SSE.
|
|
| Jul 14, 2020 at 13:50 | history | closed | whuber♦ | Needs details or clarity | |
| Jul 14, 2020 at 13:49 | history | edited | Dan Chaltiel |
edited tags
|
|
| Jul 14, 2020 at 13:32 | comment | added | whuber♦ | Please write your question in a way that does not require people to understand your code or guess your intentions. In particular, please explain what the different lines in your plots represent and what your underlying data and model are. | |
| Jul 14, 2020 at 13:14 | history | asked | Dan Chaltiel | CC BY-SA 4.0 |