Finding out how a variable was computed

[Marshaun Glover]

I am a newbie to syntax, so I hope someone can understand what I am asking.
I am working with a secondary dataset and I have a variable that has been
computed in SPSS.  Is there any syntax I can use to find out exactly how the
variable was computed?



NOTE: The information contained in this message may be privileged,
confidential, or legally protected from unauthorized disclosure. If you are
NOT the intended recipient, please note that any dissemination, distribution
or copying of this communication is strictly prohibited. If you have
received this message in error, please notify the sender immediately and
destroy this message.

Please be aware that email communication can be intercepted in transmission
or misdirected. Your use of email to communicate protected health
information to us indicates that you acknowledge and accept the possible
risks associated with such communication. Please consider communicating any
sensitive information by telephone, fax or mail. If you do not wish to have
your information sent by email, please contact the sender immediately.

[Richard Ristow]

At 10:19 AM 1/22/2007, Marshaun Glover wrote:

>Is there any syntax I can use to find out exactly how the variable was
>computed?

I'm afraid the answer is, no; and there couldn't be, unless SPSS kept
the transformation syntax as well as executing it. (Jon Peck has
developed some Python code that, as I understand it, stores the
transformation code as a variable attribute as well as executing it.
But that won't help except for transformations written using that
module.)

That leaves you with making educated guesses, and checking them. The
most important guess is what variable or variables the derived variable
was computed from. If you know the meaning of the derived variable,
that usually constrains the source variables pretty narrowly.

Then compare the putative source variables with the computed variable,
for example by
. Scatterplot, if both are continuous
. Crosstab, if both are categorical (or discrete ordinal)
. MEANS or the like, if the source is continuous and the derived
variable categorical. Use MIN and MAX statistics, which may give you
the RECODE, or equivalent, that was used.

No magic. You may or may not be able to solve your problem this way,
though the odds are fairly good if you know the intended meaning of
your derived variable of all possible source variables, and the source
variables are in your dataset. If any of those don't apply, and the
dataset isn't documented, you're probably stuck. (That gives you a
right, however, to yell at the creator of the dataset, pretty much as
loudly as you please.)

-Good luck,
  Richard