How to calculate the coefficient of variation

[Jesse Jahrig]

Hello all,

Can anyone tell me how to calculate the coefficient of variation using
SPSS? I have searched the help and gone through the different menu
options without finding anything. Please let me know if there is
additional information I can provide.

-Jesse

This communication is intended for the use of the recipient to which it is addressed, and may contain confidential, personal and privileged information. Please contact the AADAC ITS Helpdesk at 780-422-2336 immediately if you are not the intended recipient of this communication, and do not copy, distribute, or take action relying on it. Any communication received in error, or subsequent reply, should be deleted from your system.

[Richard Ristow]

I'm surprised enough that nobody's answered this, that I think I must
be missing something.

At 03:18 PM 9/28/2006, Jesse Jahrig wrote:

>Can anyone tell me how to calculate the coefficient of variation using
>SPSS?

If, as I understand, you mean simply the standard deviation divided by
the mean, then if you want the CV of a single variable across cases,
AGGREGATE using functions MEAN and SD, and divide the results; if it's
multiple variables in one case, use functions MEAN and SD in the
transformation language (in COMPUTE statements), and divide the
results.

[Marta García-Granero]

Hi Richard and Jesse

RR> I'm surprised enough that nobody's answered this, that I think I must
RR> be missing something.

It looks that I din't spot that message until you replied... Sorry

RR> At 03:18 PM 9/28/2006, Jesse Jahrig wrote:

>>Can anyone tell me how to calculate the coefficient of variation using
>>SPSS?

RR> If, as I understand, you mean simply the standard deviation divided by
RR> the mean, then if you want the CV of a single variable across cases,
RR> AGGREGATE using functions MEAN and SD, and divide the results; if it's
RR> multiple variables in one case, use functions MEAN and SD in the
RR> transformation language (in COMPUTE statements), and divide the
RR> results.

My contribution to the topic (better late than never)

If you don't want the CV as a new variable, but simply as a statistic
in a pivot table, you can also use RATIO: this procedure computes
several statistics for the ratio of two variables, including
confidence intervals for the median and coefficient of variation (mean
and median centered). To trick RATIO to compute a cv or a median CI
for a SINGLE variable (not the ratio of two variables) you just have
to compute a variable filled with "1" and use it as denominator. The
result will be the CV for the ratio between your variable and 1, and
it is evident that any variable divided by 1 remains unchanged.

See this example (I have added also the 95%CI for the median, just for
didactic purposes), if you don't want it erase "CIN(95) MEDIAN":

DATA LIST FREE/dietgroup iron (2 F8).
BEGIN DATA
0  83 0  76 0 112 0 113 0 115 0  91 0 101 0  84 0  92 0  73
0  91 0  95 0 107 0 118 0 123 0  96 0  84 0 118 0 122 0  78
0 117 0 103 0  98 0 111 0 130 0  79 0 130 0 126 0 141 0 115
1  75 1 122 1 128 1  88 1 112 1 110 1 106 1 130 1  79 1 131
1  97 1 125 1 102 1  71 1 111 1 132 1 106 1  92 1 108 1  87
1  86 1  87 1  98 1 132 1  94 1  92 1 121 1  86 1 122 1 100
1 108 1  96 1 115 1 114 1  75 1  84 1  98 1  95 1  86 1  78
1 114 1 114 1 101 1  95 1  99 1  96 1 108 1  86 1  97 1 114
END DATA.
VAR LABEL dietgroup  'Diet Group'/
          iron 'Seric Iron levels (µg/100 ml) '.
VALUE LABEL dietgroup 0 'Standard' 1 'Vegetarian'.

TEMPORARY.
COMPUTE one=1.
RATIO STATISTICS iron  WITH one  BY dietgroup
  /PRINT =  CIN(95) MEDIAN MNCOV .



--
HTH & Regards,
Dr. Marta García-Granero,PhD           mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)