Missing cases ... or are they?

[Matthew Reeder]

Hey list,
 
Quick question. I have v14 installed my work computer. Several other coworkers also use SPSS, though different versions (13 and 16). I have a question about something that happened recently that's a bit disconcerting. I've not seen anyone else mention this before, so I'm curious.
 
Long story short ... I sent an e-mail the other day to one co-worker (using v13) expressing my complete confusion to her statement that a dataset that we were working with at the time contained 611 cases. Upon opening the dataset, I only saw 577 cases - not 611. Initially, it got to the point where we were taking screenshots of the dataset and e-mailing the shots back and forth to each other, as if to say, "Ha, you're wrong!"
 
However, another co-worker soon got involved and sent over a screenshot from his computer (v16). To my chagrin, he also was coming up with 611 cases (it seems I've lost the argument). Hmm. So, I went back to my dataset and tried running a few commands to see what happened. First, I ran frequencies on a categorical variable in the dataset ... total number of cases reported was 577. Then, I tried running descriptives on a couple variables in the dataset and, sure enough, it started reporting results for 611 cases, not 577. There was no intervening command run between the frequencies and the descriptives that would change the number of cases for which results were reported. After this, I tried running a few more procedures, all of which reported results for 611 cases (including the inital frequencies that I ran that reported results for 577 cases).
 
My question: What is going on that it would do this? If this were to be a bug, my concern would stem from the realization that I've been using this version for a couple of years now, and would wonder if any other analyses I've been running have been affected by this.
 
 
 
Thanks,
Matt




====================To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[ViAnn Beadle]

I seem to recall some defect on the number of cases visible within the data
editor in 14.0 that was fixed in an update. You don't see the cases but they
are there--it is just a problem within the data editor and not within the
actual data file.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Matthew Reeder
Sent: Wednesday, July 09, 2008 3:57 PM
To: [hidden email]
Subject: Missing cases ... or are they?

Hey list,

Quick question. I have v14 installed my work computer. Several other
coworkers also use SPSS, though different versions (13 and 16). I have a
question about something that happened recently that's a bit disconcerting.
I've not seen anyone else mention this before, so I'm curious.

Long story short ... I sent an e-mail the other day to one co-worker (using
v13) expressing my complete confusion to her statement that a dataset that
we were working with at the time contained 611 cases. Upon opening the
dataset, I only saw 577 cases - not 611. Initially, it got to the point
where we were taking screenshots of the dataset and e-mailing the shots back
and forth to each other, as if to say, "Ha, you're wrong!"

However, another co-worker soon got involved and sent over a screenshot from
his computer (v16). To my chagrin, he also was coming up with 611 cases (it
seems I've lost the argument). Hmm. So, I went back to my dataset and tried
running a few commands to see what happened. First, I ran frequencies on a
categorical variable in the dataset ... total number of cases reported
was 577. Then, I tried running descriptives on a couple variables in the
dataset and, sure enough, it started reporting results for 611 cases, not
577. There was no intervening command run between the frequencies and the
descriptives that would change the number of cases for which results were
reported. After this, I tried running a few more procedures, all of which
reported results for 611 cases (including the inital frequencies that I ran
that reported results for 577 cases).

My question: What is going on that it would do this? If this were to be a
bug, my concern would stem from the realization that I've been using this
version for a couple of years now, and would wonder if any other analyses
I've been running have been affected by this.



Thanks,
Matt




=======
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD