Bulk Change of Value Labels for all Variables

[tracker925]

I am a Noob with SPSS for Windows.

I have  233 variables for which I need to add value labels.

There are only 30 or so different values for the entire data set.  (1=yes, 2= no, 3=maybe..30=etc).    Some of the Variables only have say values 1 and 2;  some variables have values  3,4,5.   If I do one mass labeling (add values 1-30 to all variables) will this mess me up on analysis or  will values  that do not appear as a response to a variable just be absent in say frequency or crosstabulation result?  

If the latter then I think I can just do the one mass labeling.  How do I do one mass labeling?  

Could you also tell me an alternative whereby I pick say 100 variables and do the value labels for just those 100.  

If coding is involved (outside the windows ui please lead me sloowly through it.  Thanks.

Ted

[Albert-Jan Roskam]

Hello,

You can simply use the same value labels for all vars, even if the values do not actually occur. If the variables are contiguous, you can use TO to save typing:
value labels v1 to v10 v15 to v100 1 'yes' 2 'no' 3 'maybe' 9 'no answer'.

Cheers!!
Albert-Jan

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
All right, but apart from the sanitation, the medicine, education, wine, public order, irrigation, roads, a fresh water system, and public health, what have the Romans ever done for us?
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

--- On Wed, 7/13/11, tracker925 <[hidden email]> wrote:

From: tracker925 <[hidden email]>
Subject: [SPSSX-L] Bulk Change of Value Labels for all Variables
To: [hidden email]
Date: Wednesday, July 13, 2011, 10:23 PM

I am a Noob with SPSS for Windows.

I have  233 variables for which I need to add value labels.

There are only 30 or so different values for the entire data set.  (1=yes,
2= no, 3=maybe..30=etc).    Some of the Variables only have say values 1 and
2;  some variables have values  3,4,5.   If I do one mass labeling (add
values 1-30 to all variables) will this mess me up on analysis or  will
values  that do not appear as a response to a variable just be absent in say
frequency or crosstabulation result?

If the latter then I think I can just do the one mass labeling.  How do I do
one mass labeling?

Could you also tell me an alternative whereby I pick say 100 variables and
do the value labels for just those 100.

If coding is involved (outside the windows ui please lead me sloowly through
it.  Thanks.

Ted









--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Bulk-Change-of-Value-Labels-for-all-Variables-tp4584232p4584232.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@... (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

[David Marso]

Please consult the documentation for the VALUE LABELS command.
You will be doing yourself a big favor by skimming the manual and getting acquainted with the functionality!
"If coding is involved (outside the windows ui please lead me sloowly through it.  Thanks."
Sloowly????
Is this slow enough?
It is really quite easy to do this in syntax and a royal pain in the @$$ to do it in the GUI.
VALUE LABELS <varlist> value "label" value "label" .....
eg
VALUE LABELS var1 TO var100 1 "value1" 2 "value2" ..........
VALUE LABELS var1 var2 var3 var10 TO var10 1 "value1" 2 "value2" 3 "value3" .

tracker925 wrote

I am a Noob with SPSS for Windows.

I have  233 variables for which I need to add value labels.

There are only 30 or so different values for the entire data set.  (1=yes, 2= no, 3=maybe..30=etc).    Some of the Variables only have say values 1 and 2;  some variables have values  3,4,5.   If I do one mass labeling (add values 1-30 to all variables) will this mess me up on analysis or  will values  that do not appear as a response to a variable just be absent in say frequency or crosstabulation result?  

If the latter then I think I can just do the one mass labeling.  How do I do one mass labeling?  

Could you also tell me an alternative whereby I pick say 100 variables and do the value labels for just those 100.  

If coding is involved (outside the windows ui please lead me sloowly through it.  Thanks.

Ted

[Bruce Weaver]

This could be done in one VALUE LABELS command with the various groups of variables separated by slashes; but I think there will be less typing if you do it with one VALUE LABELS command that assigns the labels for 1 and 2 to all 233 variables, followed by a series of ADD VALUE LABELS commands for the other groups of variables.  (You can read about the distinction between VALUE LABELS and ADD VALUE LABELS in the Help files.)  E.g., write some syntax similar to the following (i.e., File - New- Syntax, then start typing), and run it.

* All variables get value labels for values of 1 and 2 .
VALUE LABELS v1 to v233
 1 "Yes"
 2 "No" .

* Suppose variables v1 to v100 also have labels for 3-5.
* ADD VALUE LABELS will not overwrite the labels for 1 and 2.
ADD VALUE LABELS  v1 to v100
 3 "Maybe"
 4 "Almost certainly maybe"
 5 "Definitely maybe" .

* Some other set of variables has labels for 6-7.
ADD VALUE LABELS v110 v120 v130 v140 v150
 6 "Label for 6"
 7 "Label for 7" .

etc.

The use of the key word TO (e.g., v1 to v233) only works when the variables you want are contiguous in the file.  If they are not, you will have to list the variables individually, as in the last ADD VALUE LABELS command above.

HTH.

tracker925 wrote

I am a Noob with SPSS for Windows.

I have  233 variables for which I need to add value labels.

There are only 30 or so different values for the entire data set.  (1=yes, 2= no, 3=maybe..30=etc).    Some of the Variables only have say values 1 and 2;  some variables have values  3,4,5.   If I do one mass labeling (add values 1-30 to all variables) will this mess me up on analysis or  will values  that do not appear as a response to a variable just be absent in say frequency or crosstabulation result?  

If the latter then I think I can just do the one mass labeling.  How do I do one mass labeling?  

Could you also tell me an alternative whereby I pick say 100 variables and do the value labels for just those 100.  

If coding is involved (outside the windows ui please lead me sloowly through it.  Thanks.

Ted

[tracker925]

Thank you very much!


On 7/13/2011 5:14 PM, Bruce Weaver [via SPSSX Discussion] wrote:

This could be done in one VALUE LABELS command with the various groups of variables separated by slashes; but I think there will be less typing if you do it with one VALUE LABELS command that assigns the labels for 1 and 2 to all 233 variables, followed by a series of ADD VALUE LABELS commands for the other groups of variables.  (You can read about the distinction between VALUE LABELS and ADD VALUE LABELS in the Help files.)  E.g., write some syntax similar to the following (i.e., File - New- Syntax, then start typing), and run it.

* All variables get value labels for values of 1 and 2 .
VALUE LABELS v1 to v233
 1 "Yes"
 2 "No" .

* Suppose variables v1 to v100 also have labels for 3-5.
* ADD VALUE LABELS will not overwrite the labels for 1 and 2.
ADD VALUE LABELS  v1 to v100
 3 "Maybe"
 4 "Almost certainly maybe"
 5 "Definitely maybe" .

* Some other set of variables has labels for 6-7.
ADD VALUE LABELS v110 v120 v130 v140 v150
 6 "Label for 6"
 7 "Label for 7" .

etc.

The use of the key word TO (e.g., v1 to v233) only works when the variables you want are contiguous in the file.  If they are not, you will have to list the variables individually, as in the last ADD VALUE LABELS command above.

HTH.

tracker925 wrote:

I am a Noob with SPSS for Windows.

I have  233 variables for which I need to add value labels.

There are only 30 or so different values for the entire data set.  (1=yes, 2= no, 3=maybe..30=etc).    Some of the Variables only have say values 1 and 2;  some variables have values  3,4,5.   If I do one mass labeling (add values 1-30 to all variables) will this mess me up on analysis or  will values  that do not appear as a response to a variable just be absent in say frequency or crosstabulation result?  

If the latter then I think I can just do the one mass labeling.  How do I do one mass labeling?  

Could you also tell me an alternative whereby I pick say 100 variables and do the value labels for just those 100.  

If coding is involved (outside the windows ui please lead me sloowly through it.  Thanks.

Ted

---

If you reply to this email, your message will be added to the discussion below:

http://spssx-discussion.1045642.n5.nabble.com/Bulk-Change-of-Value-Labels-for-all-Variables-tp4584232p4584387.html

To unsubscribe from Bulk Change of Value Labels for all Variables, click here.

[John F Hall]

Bruce, Albert-Jan and David have already explained how to do it.  You're far
better using syntax than the GUI for this.  See the SPSS tutorials on my
website, especially section 1.4:  "Completing your data dictionary".  Being
a "noobie" you may well benefit from the earlier and later tutorials as
well.

If you've not used syntax before, open your *.sav file:

File > new > syntax

Write your SPSS commands (don't forget the full stops!)

Click Run > All


John F Hall

[hidden email]
www.surveyresearch.weebly.com





-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
tracker925
Sent: 13 July 2011 22:24
To: [hidden email]
Subject: Bulk Change of Value Labels for all Variables

I am a Noob with SPSS for Windows.

I have  233 variables for which I need to add value labels.

There are only 30 or so different values for the entire data set.  (1=yes,
2= no, 3=maybe..30=etc).    Some of the Variables only have say values 1 and
2;  some variables have values  3,4,5.   If I do one mass labeling (add
values 1-30 to all variables) will this mess me up on analysis or  will
values  that do not appear as a response to a variable just be absent in say
frequency or crosstabulation result?

If the latter then I think I can just do the one mass labeling.  How do I do
one mass labeling?

Could you also tell me an alternative whereby I pick say 100 variables and
do the value labels for just those 100.

If coding is involved (outside the windows ui please lead me sloowly through
it.  Thanks.

Ted









--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/Bulk-Change-of-Value-Labels-fo
r-all-Variables-tp4584232p4584232.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

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