calculating correlations between FILES

[Serra Koyuncu GMX]

Hi members,

First of all, I wish you a wonderful new year with health, happiness and
success!

With my limited knowledge of SPSS I came across my first problem of the year
:o(
I've got two files of test results (before-after measurement): Test1 and
test3
Most of the participants of the test did both tests, some of them only 1 and
some of them only 3.
I want to see whether there is a significant improvement of the participants
who do both tests. However, these are in separate files (1 and 3). My boss
said I should not merge them but find a way to make correlations over the
files. Is it possible? If yes, how can I do that?

Thank you very much in advance.

Best regards,
Serra

=====================
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

[Lemon, John S.]

Serra

As far as I know there is no way to do this apart from  merging the two files; I willing to be corrected by others but from 30+ years experience of using SPSS all the analysis is done on a single 'active' file and the only time SPSS will see two file is during the merge process. I suggest you do the merge.

Best Wishes

John S. Lemon

DIT - Student Liaison Officer
University of Aberdeen
Edward Wright Building: Room G51
Tel:  [phone]
Fax: [phone]



-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Serra Koyuncu GMX
Sent: 01 January 2008 10:02
To: [hidden email]
Subject: calculating correlations between FILES

Hi members,

First of all, I wish you a wonderful new year with health, happiness and
success!

With my limited knowledge of SPSS I came across my first problem of the year
:o(
I've got two files of test results (before-after measurement): Test1 and
test3
Most of the participants of the test did both tests, some of them only 1 and
some of them only 3.
I want to see whether there is a significant improvement of the participants
who do both tests. However, these are in separate files (1 and 3). My boss
said I should not merge them but find a way to make correlations over the
files. Is it possible? If yes, how can I do that?

Thank you very much in advance.

Best regards,
Serra

=====================
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

[Marta Garcia-Granero]

> Serra
>
> As far as I know there is no way to do this apart from  merging the two files; I willing to be corrected by others but from 30+ years experience of using SPSS all the analysis is done on a single 'active' file and the only time SPSS will see two file is during the merge process. I suggest you do the merge.
>

I agree, you must merge the files. I hope you have a common variable in
both files (a way of identifying pairs of cases).

Happy New Year
Marta

>
>
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Serra Koyuncu GMX
> Sent: 01 January 2008 10:02
> To: [hidden email]
> Subject: calculating correlations between FILES
>
> Hi members,
>
> First of all, I wish you a wonderful new year with health, happiness and
> success!
>
> With my limited knowledge of SPSS I came across my first problem of the year
> :o(
> I've got two files of test results (before-after measurement): Test1 and
> test3
> Most of the participants of the test did both tests, some of them only 1 and
> some of them only 3.
> I want to see whether there is a significant improvement of the participants
> who do both tests. However, these are in separate files (1 and 3). My boss
> said I should not merge them but find a way to make correlations over the
> files. Is it possible? If yes, how can I do that?
>
> Thank you very much in advance.
>
> Best regards,
> Serra
>
> =====================
> 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
>
>

=====================
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

[Dennis Deck]

 To calculate a correlation you must join the files but this can be done
temporarily (ie, only for that job), linking on the respondent ID using
MATCH FILES.

Something like following (assuming
  a) ID represents the participant identifier,
  b) both files were first sorted by ID (and checked for duplicates)
  c) file1/file3 represent the file names in quotes):

MATCH FILES
  File= "file1" /In=Pre
    /Rename= (Item1=Item10)
 /File= "file3" /In=Post
    /Rename= (Item1=Item11)
   /By= ID /Map  .

CROSSTABS Tables= Pre by Post .
CORRELATIONS Vars= Item10 Item11 .


Note that you probably need to rename the variables you want from each
file
as part of this command.  In my example I renamed Item1 as Item10 in
File1 and Item11 in File3.
The /In= creates a variable with 1 to indicate the case had that record
or a 0 if not.
The CROSSTAB will show you how many had both waves or a single wave.


Note that as a rule, I recommend always working from the raw data. I
typically do any further manipulations and calculations (like joining
the two files) as part the current job.  In that way if you get an
update (or find a programming mistake), you simply rerun.

-----Original Message-----
From: Serra Koyuncu GMX [mailto:[hidden email]]
Sent: Tuesday, January 01, 2008 2:02 AM
Subject: calculating correlations between FILES

Hi members,

First of all, I wish you a wonderful new year with health, happiness and
success!

With my limited knowledge of SPSS I came across my first problem of the
year :o( I've got two files of test results (before-after measurement):
Test1 and
test3
Most of the participants of the test did both tests, some of them only 1
and some of them only 3.
I want to see whether there is a significant improvement of the
participants who do both tests. However, these are in separate files (1
and 3). My boss said I should not merge them but find a way to make
correlations over the files. Is it possible? If yes, how can I do that?

Thank you very much in advance.

Best regards,
Serra

=====================
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