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

# Re: st: Interpreting summarize, detail

 From Syed Basher To statalist@hsphsun2.harvard.edu Subject Re: st: Interpreting summarize, detail Date Sat, 5 Feb 2011 23:02:42 -0800 (PST)

```Thank you Phil, and really sorry for not following Stata FAQ 2.2. Will keep this
in mind for future posting.

Syed Basher

----- Original Message ----
From: Phil Clayton <philclayton@internode.on.net>
To: statalist@hsphsun2.harvard.edu
Sent: Sun, February 6, 2011 9:19:20 AM
Subject: Re: st: Interpreting summarize, detail

You need to read it in columns, not rows. There are 3 columns of results -
percentiles, smallest and largest 4 values, and other stats.

So the smallest 4 values for mpg are 12, 12, 14 and 14; the largest 4 values for
mpg are 34, 35, 35 and 41. These are completely separate from the percentiles
which are in the first column of results.

by replying to a previous posting"

Phil

On 06/02/2011, at 5:09 PM, Syed Basher wrote:

> Hello all,
>
> I am having some difficulty in interpreting the basic summary statistics.
> Consider the following:
>
> . sysuse auto
> (1978 Automobile Data)
>
> . summarize mpg, detail
>
>                      Mileage (mpg)
> -------------------------------------------------------------
>    Percentiles      Smallest
> 1%            12            12
> 5%            14            12
> 10%          14            14      Obs                      74
> 25%          18            14      Sum of Wgt.          74
>
> 50%          20                          Mean            21.2973
>                            Largest      Std. Dev.      5.785503
> 75%          25            34
> 90%          29            35      Variance      33.47205
> 95%          34            35      Skewness    .9487176
> 99%          41            41      Kurtosis      3.975005
>
> In the above output, the largest value in 75% is 34, while the starting value
>of
>
> 90% is 29. The same is with the 95%. Why this is so? Shouldn't the percentile
> value be monotonically increasing? I am interested in this because in one of my
>
> own data, I have obtained the following output:
>
>                          uprice
> -------------------------------------------------------------
>    Percentiles              Smallest
> 1%    .0022779        .0001697
> 5%      .01875          .0002087
> 10%    .0581161      .0010804      Obs                626
> 25%    .3826962      .0012382      Sum of Wgt.    626
>
> 50%    1.667209                                Mean          79.44026
>                                  Largest            Std. Dev.      432.1083
> 75%    9.730152      2658.562
> 90%    77.09222      3077.629      Variance      186717.6
> 95%    363.5599      5423.877      Skewness      11.04796
> 99%      1490.65      7004.734      Kurtosis      151.0011
>
> where as you can see I have a similar problem (largest value in 75% is much
> higher than the starting value in 90% and so on). I am guessing that this is
>due
>
> to the 3rd (skewness) and 4th (kurtosis) moments of the distribution. But I do

> not have a convincing interpretation/explanation. Your help will be much
> appreciated.
>
> Regards,
>
> Syed Basher
> Qatar National Food Security Programme
>
>
>
>
> *
> *  For searches and help try:
> *  http://www.stata.com/help.cgi?search
> *  http://www.stata.com/support/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/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/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```