Deborah L. Rhodes
> My first question: Does categorizing numerical data into
> quartiles require
> two steps or one? The original data(totdolars)is in
> dollar amounts.
>
> I have tried to go directly to the quartiles: (using both g
> and as noted below)
> xtile totdolarsqtl=totdolars, nquantiles(25);
>
> I have the same problem if I try pctile to create
> percentiles to bring to
> quartiles. It seems the syntax should be similar.
> pctile totdolarspc=totdolars, nquantiles(1);
>
> I have substituted totdolars (my variable with original
> dollar amount data)
> for exp, which according to the help file for "creating
> variables using
> percentiles" is "typically just another variable."
You are misreading the -nquantiles()- option. Here is
what the on-line help says:
nquantiles(#) specifies the number of quantiles.
The command computes percentiles corresponding to
percentages 100*k/# for k=1, 2, ..., #-1. For example,
nquantiles(10) requests that the 10th, 20th, ..., 90th
percentiles be computed. The default is nquantiles(2);
i.e., the median is computed.
For quartiles specify -nq(4)-:
. xtile totdolarsqtl = totdolars, nq(4)
The result is a variable with values 1 to 4, depending
on where values lie relative to the lower quartile,
median and upper quartile. All values specified will
be categorised.
. pctile totdolarspc = totdolars, nq(4)
The result is a variable with just 3 [sic] non-missing
values, containing the lower quartile, median and upper
quartile. That is probably not what you want.
Nick
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
