Timeline for Are estimates of regression coefficients uncorrelated?
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 22, 2018 at 14:51 | comment | added | whuber♦ | @Heisenberg I am grateful for your perceptive observation: it enabled me to correct a substantial error in the discussion of the multiple regression case. | |
| Feb 22, 2018 at 14:50 | history | edited | whuber♦ | CC BY-SA 3.0 |
added 447 characters in body
|
| Feb 21, 2018 at 23:07 | comment | added | whuber♦ | @Heisenberg That's a good point. I was unclear about this. There's no ambiguity in the case of two columns, but I need to think how to improve the presentation for the case of more columns. | |
| Feb 21, 2018 at 22:47 | comment | added | Heisenberg | The answer says "[...] off-diagonal elements, which are just the dot products of the two columns of X." This is true for $X'X$, not $(X'X)^{-1}$ however? | |
| Jun 25, 2014 at 18:30 | vote | accept | arnab | ||
| Jun 25, 2014 at 14:43 | history | answered | whuber♦ | CC BY-SA 3.0 |