Timeline for Why in Box-Cox method we try to make x and y normally distributed, but that's not an assumption for linear regression?
Current License: CC BY-SA 3.0
22 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Jul 9 at 9:00 | history | protected | kjetil b halvorsen♦ | ||
Jul 9 at 6:39 | answer | added | Gene Uniana | timeline score: 1 | |
Feb 5 at 12:50 | history | edited | kjetil b halvorsen♦ |
edited tags
|
|
Jan 4, 2019 at 6:35 | vote | accept | Andre | ||
Oct 30, 2017 at 2:49 | vote | accept | Andre | ||
Jan 4, 2019 at 6:35 | |||||
Oct 28, 2017 at 6:16 | vote | accept | Andre | ||
Oct 28, 2017 at 6:16 | |||||
Oct 28, 2017 at 6:16 | vote | accept | Andre | ||
Oct 28, 2017 at 6:16 | |||||
Oct 27, 2017 at 19:32 | comment | added | Andre | @whuber or ask questions on Cross Validated :) | |
Oct 27, 2017 at 19:30 | answer | added | Andre | timeline score: 1 | |
Oct 27, 2017 at 16:46 | comment | added | whuber♦ |
@Matthew It's a losing battle: new books on some version of R plus statistics are appearing faster than any one person could even read them. A better response is to write a better book--but that's obviously a significant effort!
|
|
Oct 27, 2017 at 16:01 | comment | added | Matthew Drury | @whuber That "since then" is particularly bad. Have you ever considered writing the authors of these texts? | |
Oct 27, 2017 at 15:41 | answer | added | Aksakal | timeline score: 3 | |
Oct 26, 2017 at 16:31 | comment | added | kjetil b halvorsen♦ | @whuber: thanks for this references, I will chech them before extending my answer. | |
Oct 26, 2017 at 16:16 | comment | added | Andre | @whuber okay, I see! Thanks so much for clarification. | |
Oct 26, 2017 at 16:16 | comment | added | whuber♦ | I'm not making this up, either: see John Tukey's book EDA (Addison-Wesley 1977) or Hoaglin, Mosteller, & Tukey, Understanding Robust and Exploratory Data Analysis (J. Wiley 1983). | |
Oct 26, 2017 at 16:13 | comment | added | whuber♦ | Let me be perfectly clear: the purpose of the Box-Cox transformation is not to make data look as normal as possible, nor is it necessary to calculate it with that aim in mind. Moreover, the aims of (1) achieving a symmetric (or Normal) distribution of residuals, (2) linearizing a relationship, and (3) achieving constant conditional variance are distinctly different and will not necessarily be achieved by the same (or any) Box-Cox transformation. The quotation is a narrow, specialized approach to finding a Box-Cox transformation and its inference ("since then...") is flatly incorrect. | |
Oct 26, 2017 at 15:38 | comment | added | Andre | @whuber I agree that box-cox method has so many uses, but what I’m also confused is in linear regression we don’t assume x and y to be normally distributed while what box-cox transformation calculates is actually to make them as normal as possible. And the quotation explains that this is because then we are more confident that x and y have linear relationship. But if so, why don’t we assume this for linear regression? | |
Oct 26, 2017 at 15:22 | history | tweeted | twitter.com/StackStats/status/923570489192271873 | ||
Oct 26, 2017 at 14:40 | comment | added | whuber♦ | Neither the quotation nor the preceding comment are fully general. Although the Box-Cox transformation can be implemented with the aims given in the quotation and using an ML method alluded to by @Kjetil, yet it's far more general than that: (1) it can be used to symmetrize distributions and/or create near-constant variances and/or linearize relationships; (2) it should be estimated using robust exploratory methods rather than the much more limited parametric maximum likelihood methods offered in most software packages. See stats.stackexchange.com/a/3530, for instance. | |
Oct 26, 2017 at 11:00 | answer | added | kjetil b halvorsen♦ | timeline score: 6 | |
Oct 26, 2017 at 6:28 | comment | added | kjetil b halvorsen♦ | In reality, box-cox transformation finds a transformation that homogenizes variance, and constant variance is an assumption! The crux of the matter is that boxcox uses a constant-variance normal likelihood. | |
Oct 26, 2017 at 5:50 | history | asked | Andre | CC BY-SA 3.0 |