Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: composite index


From   "Joon G. Park" <calgarypark@politics.tamu.edu>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   st: composite index
Date   Thu, 16 Nov 2006 19:39:44 -0600

Hi,

 

I have a quick question on creating a composite index. I've searched for this topic on Stata listserv, but without success.

 

But, anyways, here is my situation:

 

I have the following three variables:  

 

Variable "A" (Scale: millions; range: from hundreds of thousands to millions; Unit: dollar)

 

Variable "B" (Scale: hundreds; Range: from single digits to three double digits; Unit: "size")

 

Variable "C" (Scale: hundreds; Range: from less than ten to tens of thousands; Unit: personnel)

 

 

I would like to have equal weights attached for each variable that will comprise the composite index (I may use weighted index later on, depending on how my theory works out). This would seem accomplished easily enough via "generate" command with relevant math syntax.

 

Although the data is noticeably non-normal, this may be remedied via some type of log transformation. And, the issue with the different scales may also be addressed via requisite transformation.

 

But the problem with differences in units still seems to pose a problem. Can anyone recommend some way out of this, so that an index can be constructed? Of course, corollary advice on the command structure would also be helpful.

 

An additional note: I am not sure if a factor analysis approach would also work, since a cursory exam of the data does not suggest a high level of positive correlation between variable C and the other two variables (although there does seem to be a high positive correlation between variables B and C). I am also reluctant to go with just a two-indicator factor.

 

Any feedback would be greatly appreciated. Much thanks.

 

 

Joon.

 

 




© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index