Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Transformation of variable with pos/neg values via asinh


From   Nick Cox <[email protected]>
To   "[email protected]" <[email protected]>
Subject   Re: st: Transformation of variable with pos/neg values via asinh
Date   Tue, 19 Nov 2013 10:40:59 +0000

David's suggestion that with a balance, namely

balance = gain - loss

you are better off with

transform of gain - transform of loss

than with with

transform of (gain - loss)

echoes John Tukey's advice in Exploratory data analysis
(Addison-Wesley, Reading, MA, 1977). Tukey didn't give an example, and
I don't recall ever seeing one in his work.

However, this is a counsel of perfection, as often the individual gain
and loss values are not accessible.

Like Austin I've been advocating cube roots for some while. I often
work with audiences for whom asinh is beyond esoteric; I've even
encountered considerable resistance to cube roots. Some people seem
reluctant to try anything not sanctioned by a hundred texts. But they
got a puff in

SJ-11-1 st0223  . . . . . . . . . . . . . . . . . . . Stata tip 96: Cube roots
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q1/11   SJ 11(1):149--154                                (no commands)
        tip showing the use of the cube function and cube roots

No pretence: using cube roots is distinctly ad hoc (which, in modern
idiom, I translate freely as "fit for purpose"). So is asinh.


Nick
[email protected]


On 19 November 2013 01:43, David Hoaglin <[email protected]> wrote:
> Dear Steven,
>
> The presence of extreme values (relative to the variation in the bulk
> of the data) suggests that the variable does not have a normal
> distribution.  Do you have a mixture of some sort?
>
> What is the source of the negative values of CEO wealth?
>
> A variable that has both positive and negative values is often the
> result of taking the difference between two positive quantities.  If
> some sort of transformation seems likely to be useful, it is usually a
> good idea to transform the variables that are the basis for the
> difference.
>
> Regards,
>
> David Hoaglin
>
> On Mon, Nov 18, 2013 at 6:21 PM, Steven Delbright
> <[email protected]> wrote:
>> Dear Austin:
>>
>> Thanks so much for the quick response! I indeed struggle with a
>> "wealth" dependent variable (CEO wealth to be exact). Do you have some
>> sample citations for academic work transforming wealth variables (not
>> necessarily in management but also other fields).
>>
>> Also, I found your comment regarding using a different link. I never
>> considered this option before but always transformed my DV. What link
>> would I use for untransformed wealth variables with fat tails?
>>
>> Thanks much!
>> Steven
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index