Date:         Tue, 28 Jan 2003 09:55:00 +0100
Reply-To:     Asesoría Bioestadística
              <bioestadistica@eresmas.net>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Asesoría Bioestadística
              <bioestadistica@eresmas.net>
Subject:      Re: Non SPSS Question (or yes?)
Content-Type: text/plain; charset=us-ascii

> Listers,
>
> Would anybody please help with how to calculate a coeffiecient of variation
> (CV) for Body Mass Index (BMI) [weight/length](squared) by gestational age?
>
> apologies for this non SPSS question bu i figured there could be people
> working on growth on this forum.
>
> Cathy

It is also an SPSS question. SPSS 11 can calculate coefficients of variation
(it is a bit hidden, inside RATIO):

* Sample data *.

INPUT PROGRAM.
- VECTOR x(2).
- LOOP #I = 1 TO 50.
- LOOP #J = 1 TO 2.
- COMPUTE x(#J) = 100+NORMAL(15).
- END LOOP.
- END CASE.
- END LOOP.
- END FILE.
END INPUT PROGRAM.
execute.
RANK  VARIABLES = x1
  /NTILES(4) INTO group
  /PRINT = NO.

* Compute CV with SPSS 11 *.

temporary.
COMPUTE one = 1 .
RATIO STATISTICS x2  WITH one  BY group  (ASCENDING)
  /PRINT =  MNCOV .

This method can also be used to obtain 95%CI for medians:

temporary.
COMPUTE one = 1 .
RATIO STATISTICS x2  WITH one  BY group  (ASCENDING)
  /PRINT =  CIN(95) MEDIAN .

HTH

Marta Garcia-Granero