Richard:
Yes, the syntax was correct and was running correctly. I was using it on
an inherited file where the previous analyst had, for some unknown
reason, coded the responses to a question along the lines of, "What
percentage of your widgets are green?" as one more than the percentage
entered by the respondent. Thus, a zero was recorded as a 1, etc. After
running the syntax, I ran a summary case report on the original
variables and the "fixed" variables and the report appeared as though
the no changes had occurred. I checked the actual data set and fund
that, in fact, the syntax had run as intended. Thanks to both you and
Edward for your response.
-----Original Message-----
From: Richard Ristow [mailto:
[hidden email]]
Sent: Sunday, July 02, 2006 7:12 PM
To: Moffitt, James (West);
[hidden email]
Cc: Edward Boadi
Subject: Re: Create new variable with values that are 1 less than values
of existing variable
At 11:25 AM 6/30/2006, Jim Moffitt wrote:
>I've run the following syntax only to get the same values in my "fixed"
>variables as in the original variables?
>
>compute q66Fixed = q66-1.
>compute q67Fixed = q67-1.
>compute q68Fixed = q68-1.
>execute.
>
>What am I doing wrong?
Did you ever solve this one? Unlike Edward Boadi, I haven't run test
code; but like him, I don't see any source of trouble.
"The more obscure the observed behavior, the simpler is the underlying
problem." In this case, to reach that simplicity, we need all your code,
and perhaps test data.
Or, of course, maybe you solve it long ago.
-Good luck,
Richard
Locked
