Comments on: The 2007 e-readiness rankings: comments and critiques http://ictlogy.net/20070502-the-2007-e-readiness-rankings-comments-and-critiques/ Information Society, Digital Divide, ICT4D Sun, 05 Feb 2012 21:25:10 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: Ismael Peña-López http://ictlogy.net/20070502-the-2007-e-readiness-rankings-comments-and-critiques/comment-page-1/#comment-87383 Ismael Peña-López Wed, 03 Dec 2008 15:05:58 +0000 http://ictlogy.net/20070502-the-2007-e-readiness-rankings-comments-and-critiques/#comment-87383 One year and a half later, I have to admit that Phillippa was absolutely right and I was completely wrong. Sorry to have been so slow to see it. One year and a half later, I have to admit that Phillippa was absolutely right and I was completely wrong.

Sorry to have been so slow to see it.

]]> By: Ismael Peña-López http://ictlogy.net/20070502-the-2007-e-readiness-rankings-comments-and-critiques/comment-page-1/#comment-28785 Ismael Peña-López Mon, 16 Jul 2007 20:49:54 +0000 http://ictlogy.net/20070502-the-2007-e-readiness-rankings-comments-and-critiques/#comment-28785 Dear Phillippa, Thanks for the point, but I actually meant to use the R<sup>2</sup> in the first figure and the Beta in the second one. Of course, the explanation is not technically accurate: I just wanted to give a faint idea on how both indices try to explain the same thing. In other words: if when doing a regression among the NRI and the EUI we got R<sup>2</sup> = 1, we could then (more or less) state that they both explain exactly the same thing. I know I should be more "formal" in my explanations, but I try to keep things simple :) Dear Phillippa,

Thanks for the point, but I actually meant to use the R2 in the first figure and the Beta in the second one. Of course, the explanation is not technically accurate: I just wanted to give a faint idea on how both indices try to explain the same thing.

In other words: if when doing a regression among the NRI and the EUI we got R2 = 1, we could then (more or less) state that they both explain exactly the same thing.

I know I should be more “formal” in my explanations, but I try to keep things simple :)

]]> By: Phillippa Biggs http://ictlogy.net/20070502-the-2007-e-readiness-rankings-comments-and-critiques/comment-page-1/#comment-28720 Phillippa Biggs Mon, 16 Jul 2007 14:47:52 +0000 http://ictlogy.net/20070502-the-2007-e-readiness-rankings-comments-and-critiques/#comment-28720 I am not sure you are plotting the wrong regression coefficient! The R2 coefficient or COEFFICIENT OF DETERMINATION only ever explains the amount of variation explained by your model or fraction of the variance in yi that is accounted for by a linear fit of xi to yi (see for example http://en.wikipedia.org/wiki/Coefficient_of_determination). What you need is the CORRELATION COEFFICIENT or Beta that explains the degree of relationship between two variables (but does not imply that this relationship is causal). See http://en.wikipedia.org/wiki/Correlation_coefficient My best wishes, Phillippa Biggs I am not sure you are plotting the wrong regression coefficient! The R2 coefficient or COEFFICIENT OF DETERMINATION only ever explains the amount of variation explained by your model or fraction of the variance in yi that is accounted for by a linear fit of xi to yi (see for example http://en.wikipedia.org/wiki/Coefficient_of_determination). What you need is the CORRELATION COEFFICIENT or Beta that explains the degree of relationship between two variables (but does not imply that this relationship is causal). See http://en.wikipedia.org/wiki/Correlation_coefficient

My best wishes, Phillippa Biggs

]]>