Timeline for Meaning of "Overdispersion" in Statistics
Current License: CC BY-SA 4.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 9, 2022 at 23:56 | history | edited | kjetil b halvorsen♦ | CC BY-SA 4.0 |
deleted 11 characters in body; edited tags
|
| Dec 4, 2021 at 6:19 | comment | added | Glen_b | More generally, anything where the variance is larger than some model would suggest might be called overdispersed, by extension from the Poisson case (which is where the term originates). In a GLM context you can estimate dispersion, but it will only be called overdispersed for cases where the dispersion parameter is already specified (binomial and Poisson, most typically) | |
| Dec 4, 2021 at 6:13 | comment | added | Glen_b | With something that might be expected t be Poisson, overdispersion is specifically for the situation where the variance is greater than mean (by comparison with the Poisson, where in the population the variance is equal the mean), not just varying with it. Consider the binomial, for example, which have variance that is a function of the mean but which is always less than the mean (by a factor of $1-p$). That would not be called overdispersed relative to the Poisson -- quite the opposite. | |
| Dec 3, 2021 at 19:51 | answer | added | schmettow | timeline score: 2 | |
| Dec 3, 2021 at 10:27 | history | became hot network question | |||
| Dec 3, 2021 at 8:41 | answer | added | Ben | timeline score: 12 | |
| Dec 3, 2021 at 6:00 | history | tweeted | twitter.com/StackStats/status/1466648392436756485 | ||
| Dec 3, 2021 at 4:43 | answer | added | user318288 | timeline score: 9 | |
| Dec 3, 2021 at 2:36 | answer | added | Dave | timeline score: 22 | |
| Dec 3, 2021 at 2:27 | history | asked | stats_noob | CC BY-SA 4.0 |