Jump to content

Talk:Unit-weighted regression

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Terminology

[edit]

Is this article worthwhile and properly-named? Would it be better as "un-weighted regression" or "ordinary regression"? Melcombe (talk) 15:19, 7 April 2008 (UTC)[reply]

Weight Loss Example

[edit]

Just... Why? Who wrote that? It looks like troll work. A better introductory example is needed, preferably one without needing to resort to using an alternative definition of the word, "weight". This could obviously confuse someone with no knowledge of weighted regression, perhaps someone thinking of using, say Wikipedia, to learn about it. So this weight-loss example... Think up a better one. And flag the troll who wrote it. Please. — Preceding unsigned comment added by 130.234.68.129 (talk) 20:15, 3 June 2011 (UTC)[reply]

Reference to technical term?

[edit]

The expression "bouncing betas" occurs in the example. What does this mean? Is there a reference; wiki or otherwise? --85.224.69.186 (talk) 11:51, 18 November 2011 (UTC)[reply]

Perhaps it's from the referenced paper. In any case, it's obviously an informal way of saying that regression coefficients differed per study, as can be expected when different studies use different predictors with naïve multiple regression. QVVERTYVS (hm?) 18:39, 27 January 2014 (UTC)[reply]