Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: Quantile regression: determine to which quantile an individual belongs


From   Nick Cox <njcoxstata@gmail.com>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: Quantile regression: determine to which quantile an individual belongs
Date   Thu, 18 Apr 2013 12:25:19 +0100

We agree on the main point.

But -round(whatever, 0.01)- is not an especially good idea because
most multiples of 0.01 can not be held exactly in binary and people
who don't understand that get into awkward small messes.

A more systematic approach is explained at

FAQ     . . . . . . . . . . Calculating percentile ranks or plotting positions
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        7/02    How can I calculate percentile ranks?
                How can I calculate plotting positions?
http://www.stata.com/support/faqs/statistics/percentile-ranks-and-plotting-positions/

Nick
njcoxstata@gmail.com


On 18 April 2013 12:20, Alex Olssen <alex.olssen@gmail.com> wrote:
> I also am unclear on what you wish to achieve.  You said that you wanted to
> know the quantile of each person's residual.  I can suggest a way to
> calculate this, but I am unsure why you want it.
>
> sysuse auto, clear
> reg price length
> predict res, res
> sort res
> gen quantile = round(_n/_N, 0.01)


On 18/04/13 9:12 PM, Nick Cox wrote:

>> I can't see that "the quantile each individual belongs to" is a
>> well-defined concept.
>>
>> Clearly you  can work out a percentile rank for each response, without
>> regard to the predictors. Or you can see where individual residuals
>> lie in the distribution of residuals.
>>
>> But I don't think that either is what you are seeking.
>>
>> I am not clear whether you are thinking of a point or an interval, but
>> that's secondary.
>>
>> Stripping it down to a minimal example: We have a quantile regression
>> for weight versus height. I am a data point with a certain weight and
>> height. What quantile do I belong to? What's your definition?

On 18 April 2013 11:50, Maria Juul Hansen <maria@lindely.dk> wrote:

>>> Thank you for your comment and reference!
>>> I am aware of the endogeneity problem. However, the purpose is not to
>>> establish causal effects, just to control for the variables of interest.
>>>
>>> Do you have any recommendations regarding the problem of identifying the
>>> quantiles each individual belongs to?
*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index