Timeline for Margins in "R" throws confidence intervals that do not contain cero, yet p-value > 0.05. Is this possible?
Current License: CC BY-SA 4.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 22, 2018 at 13:15 | vote | accept | J.M. Molina | ||
| Oct 19, 2018 at 15:05 | comment | added | mdewey | I think asking this question on R-help might be productive. | |
| Oct 19, 2018 at 15:04 | comment | added | J.M. Molina | @a_statistician The package's author has a book on complex survey samples on which the procedures are based on, I should check that out. | |
| Oct 19, 2018 at 15:00 | comment | added | J.M. Molina | @St.Inkbug According to this help page in the documentation), the default seems to be 95%, but for some reason the intervals are different that those shown there for Stata. | |
| Oct 19, 2018 at 14:56 | comment | added | user158565 | @mdewey I do not use R. But I hope I could find the theory that support svyglm. I am very disappointed by the search results. | |
| Oct 19, 2018 at 14:51 | comment | added | mdewey | @a_statistician in fact the confidence intervals come from the margins package but that is not explicit about how it estimates them either. I think the OP has to look at the code for enlightenment. | |
| Oct 19, 2018 at 14:39 | answer | added | Heteroskedastic Jim | timeline score: 6 | |
| Oct 19, 2018 at 14:28 | comment | added | user158565 | I googled the information about svyglm. Its manual does not say what methods are used. So it just likes black box. I will not use it before I get the statistical theory behind the black box. | |
| Oct 19, 2018 at 5:28 | comment | added | St. Inkbug | Are you sure that the reported confidence intervals are 95%? | |
| Oct 18, 2018 at 22:02 | comment | added | Heteroskedastic Jim | The confidence intervals are not symmetric about the AME. When this is the case, this result is possible. The CI is not simply $AME \pm c \times SE$. So they are arrived at differently. The p-values are based on the Wald z-statistic, the CI are not Wald CI. Different assumptions, different results. This is assuming the package is not making a programming error. | |
| Oct 18, 2018 at 21:50 | review | First posts | |||
| Oct 19, 2018 at 0:26 | |||||
| Oct 18, 2018 at 21:46 | history | asked | J.M. Molina | CC BY-SA 4.0 |