Hi

On Tue, 5 Mar 2002, Claudia Stanny wrote:
> Does SPSS have a mechanism by which I can verify the accuracy of a large
> SPSS data file by creating two independent versions of the file and
> comparing them?

I think it would help if you explained exactly what you mean by
"creating two independent versions of the file."  Do you mean:
(1) Having the data entered twice?  (2) Entering the data once,
and then making a copy of the file electronically?  (3) Something
else?  There would seem to be no reason to expect differences for
(2).  So presumably (1) or (3) are what you mean?

>  We plan to compute and compare descriptive statistics
> computed for each variable for the two independent versions of the data,
> but this will only those identify variables with data discrepancies.  Is
> there a procedure for doing a cell-by-cell parity comparison and identify
> cells in the two files that have different entries?  I am looking for a
> procedure that will do what keypunch operators used to do -- verifying the
> accuracy of punched cards by repunching them.

SPSS does have a DO REPEAT command that allows users to do things
like compare a list of variables.  It would perhaps work
something like this:

DO REPEAT v1 = a1 TO a30  /v2 = b1 TO b30 /r = r1 TO r30
COMPUTE r = v1 EQ v2
END REPEAT

Examining variables r1 to r30 will show discrepancies between
a1-b1, a2-b2, and so on (values will be 0 for non-matches and 1
for matches).

To do this, you will also need to get SPSS to read both files and
match the corresponding records to make a single working file
before making the comparisons.

There is also an SPSS newsgroup that is VERY helpful with such
problems.  Its name is:

comp.soft-sys.stat.spss 

Best wishes
Jim

============================================================================
James M. Clark                          (204) 786-9757
Department of Psychology                (204) 774-4134 Fax
University of Winnipeg                  4L05D
Winnipeg, Manitoba  R3B 2E9             [EMAIL PROTECTED]
CANADA                                  http://www.uwinnipeg.ca/~clark
============================================================================



---
You are currently subscribed to tips as: [EMAIL PROTECTED]
To unsubscribe send a blank email to [EMAIL PROTECTED]

Reply via email to