Timeline for How to eliminate the influence of heteroscedasticity
Current License: CC BY-SA 4.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 5, 2019 at 9:00 | history | tweeted | twitter.com/StackStats/status/1191641458740256769 | ||
| Nov 4, 2019 at 23:13 | answer | added | Manuel | timeline score: 4 | |
| Nov 3, 2019 at 22:54 | comment | added | Aeeh | Thanks, guys. I just want to fit a model that could show a more precise negative relationship. LIke $x_2>x_1$ then $y_2<y_1$, but this heteroscedasticity makes the result not reliable, so I want to eliminate the influence of heteroscedasticity and make an approximately linear model. | |
| Nov 3, 2019 at 20:36 | comment | added | André.B | Also to clarify: A linear model does not mean that there needs to be a straight line relationship between x and y. The "linear" part in the name is eluding to the linear combination of predictors. Linear models include models with polynomial terms. | |
| Nov 3, 2019 at 20:34 | comment | added | André.B | You could use a generalised least squares model which allows for heteroscedasticity. | |
| Nov 3, 2019 at 20:30 | comment | added | Alexandre C-L | Just to know, why does heteroscedasticity bother you ? (Because a linear relationship seems fine in your case, except that you have to account for heteroscedasticity for the computation of standard errors) | |
| Nov 3, 2019 at 19:14 | history | edited | Aeeh | CC BY-SA 4.0 |
added 53 characters in body
|
| Nov 3, 2019 at 19:08 | history | edited | Aeeh | CC BY-SA 4.0 |
added 53 characters in body
|
| Nov 3, 2019 at 19:05 | review | First posts | |||
| Nov 3, 2019 at 20:34 | |||||
| Nov 3, 2019 at 19:02 | history | asked | Aeeh | CC BY-SA 4.0 |