SPSSX Discussion

Column widths in PASW 18

Classic

List

Threaded

10 messages Options

John F Hall

Column widths in PASW 18

I like to adjust column widths in the Data Edotor so I can widen var labs and narrow everything else. Sometimes it works, sometimes it doesn't. It's irritating to widen or narrow one column and then all the others jump to the same width. Also when I've left a Data Editor adjusted, next time I open it , it's back to the default equal width again.

a: What's the problem?

b: Why can't it save my last settings?

Mike Pritchard

Creating missing variables without overwriting existing variables

I’m working with some sparse data from an online survey, where the fields were only present for options that had responses. In addition to analysis with SPSS, I’m creating tables with Wincross. To make it easier/consistent, I’d like to have variables including all of the empty data.

I think there should be an easy way to do this within SPSS without manually creating a list. The variable names are of the form Q37_1_1, Q37_1_2, etc., Q37_2_1, Q37_2_2, etc., with each series having different missing names. The best idea I’ve had so far is to use Excel to create the list and then merge variables, but I wonder if there is an easier way using Syntax? The problem I see with COMPUTE is that it would erase data. I’m on SPSS V16

Thanks

Mike

_________________________________________________________________________
Mike Pritchard | [hidden email] | 5 Circles Research | 425-444-3410
Helping companies delight their customers to increase profits

Mike Pritchard

Re: Creating missing variables without overwriting existing variables

In reply to this post by John F Hall

So much for the merge with the file created in Excel. Maybe this is a bug, but more likely just my limited understanding of Merges. For some reason the Merge to add variables caused some of the original variables to be deleted. The deleted variables were in the original file, not the file with the new variables.

Back to the drawing board…

Mike

From: Mike Pritchard [mailto:[hidden email]]
Sent: Sunday, April 25, 2010 12:32 PM
To: '[hidden email]'
Subject: Creating missing variables without overwriting existing variables

Thanks

Mike

Maguin, Eugene

Re: Creating missing variables without overwriting existing variables

In reply to this post by Mike Pritchard

Mike,

It might be helpful to all but certainly to me if you were to give some sort
example to illustrate the problem. I'm sure some of the people on the list
know exactly what you are describing, perhaps because they have to deal with
it. For me, what you say doesn't quite make any sense because I've never
seen data like you describe.

Gene Maguin

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

John F Hall

Re: Creating missing variables without overwriting existing variables

In reply to this post by Mike Pritchard

I hope you kept backup copies of the files. Like Gene said, we need to see an example of your Data Editor if we are to help you.

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Sunday, April 25, 2010 9:32 PM

Subject: Creating missing variables without overwriting existing variables

Im working with some sparse data from an online survey, where the fields were only present for options that had responses. In addition to analysis with SPSS, Im creating tables with Wincross. To make it easier/consistent, Id like to have variables including all of the empty data.

I think there should be an easy way to do this within SPSS without manually creating a list. The variable names are of the form Q37_1_1, Q37_1_2, etc., Q37_2_1, Q37_2_2, etc., with each series having different missing names. The best idea Ive had so far is to use Excel to create the list and then merge variables, but I wonder if there is an easier way using Syntax? The problem I see with COMPUTE is that it would erase data. Im on SPSS V16

Thanks

Mike

_________________________________________________________________________
Mike Pritchard | [hidden email] | 5 Circles Research | 425-444-3410
Helping companies delight their customers to increase profits

Art Kendall

Re: Creating missing variables without overwriting existing variables

In reply to this post by Mike Pritchard

I am not sure I completely follow. See if the syntax below does what you want.

I created 3 excel files

id	q37_1	Q37_2	q37_3
1001	5	4	3
1002	2	1	3
1003	3	4	3

id	q38_1	Q38_2	q39_2
1001	7	12	25
1002	8	11	26
1003	9	13	27
1004	6	6	6
1005	3		3
1006		12

id	q37_1	q37_3
2001	22	51
2002	23	52
3003	24	53

GET DATA /TYPE=XLSX /FILE='C:\Users\Art\Desktop\excel1.xlsx' /SHEET=name 'Sheet1' /CELLRANGE=full /READNAMES=on /ASSUMEDSTRWIDTH=32767. DATASET NAME DataSet1 WINDOW=FRONT. list. sort cases by id. GET DATA /TYPE=XLSX /FILE='C:\Users\Art\Desktop\excel2.xlsx' /SHEET=name 'Sheet1' /CELLRANGE=full /READNAMES=on /ASSUMEDSTRWIDTH=32767. DATASET NAME DataSet2 WINDOW=FRONT. list. sort cases by id. GET DATA /TYPE=XLSX /FILE='C:\Users\Art\Desktop\excel3.xlsx' /SHEET=name 'Sheet1' /CELLRANGE=full /READNAMES=on /ASSUMEDSTRWIDTH=32767. DATASET NAME DataSet3 WINDOW=FRONT. list. sort cases by id. match file file=dataset1 /file=dataset2 /file=dataset3 /by id /map. dataset name combo. dataset activate combo. list.

Art Kendall
Social Research Consultants

On 4/25/2010 3:32 PM, Mike Pritchard wrote:

I’m working with some sparse data from an online survey, where the fields were only present for options that had responses. In addition to analysis with SPSS, I’m creating tables with Wincross. To make it easier/consistent, I’d like to have variables including all of the empty data.

I think there should be an easy way to do this within SPSS without manually creating a list. The variable names are of the form Q37_1_1, Q37_1_2, etc., Q37_2_1, Q37_2_2, etc., with each series having different missing names. The best idea I’ve had so far is to use Excel to create the list and then merge variables, but I wonder if there is an easier way using Syntax? The problem I see with COMPUTE is that it would erase data. I’m on SPSS V16

Thanks

Mike

_________________________________________________________________________
Mike Pritchard | [hidden email] | 5 Circles Research | 425-444-3410
Helping companies delight their customers to increase profits

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

Art Kendall
Social Research Consultants

Mike Pritchard

Re: Creating missing variables without overwriting existing variables

Thanks Art. I’m not sure if the problem I encountered with the merge (later message) – other variables being deleted – was related, but what I meant by sparse data was that some variables were already there from the list of 52 for each option. What I was hoping to achieve was that new variables would be added, but that existing once would remain intact. I couldn’t figure out what was causing the deletion of variables that weren’t in the matched set, so I’m in the process of adding the variables by hand.

This particular set of questions caused me other problems with the export from the survey tool. The SPSS export is normally good, but this part was useless. It wasn’t a very complex question set, but perhaps complex enough to mess things up. The first question was a setup for a list of organizations to compare against. Then subsequent questions asked about particular attributes for the organizations checked in the first question, including Other/Specifies. The SPSS export only included results for option 1. The CSV export had complete results, but only for those checked. It probably would have been worth me adding the other variables into the CSV file before the original merge. Another option would have been to change each Wincross table to only have rows for populated variables, but that seemed more error prone.

I live and (sometimes) learn.

Mike

From: Art Kendall [mailto:[hidden email]]
Sent: Sunday, April 25, 2010 2:29 PM
To: Mike Pritchard
Cc: [hidden email]
Subject: Re: [SPSSX-L] Creating missing variables without overwriting existing variables

I am not sure I completely follow. See if the syntax below does what you want.

I created 3 excel files

id	q37_1	Q37_2	q37_3
1001	5	4	3
1002	2	1	3
1003	3	4	3

id	q38_1	Q38_2	q39_2
1001	7	12	25
1002	8	11	26
1003	9	13	27
1004	6	6	6
1005	3		3
1006		12

id	q37_1	q37_3
2001	22	51
2002	23	52
3003	24	53

GET DATA /TYPE=XLSX
/FILE='C:\Users\Art\Desktop\excel1.xlsx'
/SHEET=name 'Sheet1'
/CELLRANGE=full
/READNAMES=on
/ASSUMEDSTRWIDTH=32767.
DATASET NAME DataSet1 WINDOW=FRONT.
list.
sort cases by id.
GET DATA /TYPE=XLSX
/FILE='C:\Users\Art\Desktop\excel2.xlsx'
/SHEET=name 'Sheet1'
/CELLRANGE=full
/READNAMES=on
/ASSUMEDSTRWIDTH=32767.
DATASET NAME DataSet2 WINDOW=FRONT.
list.
sort cases by id.
GET DATA /TYPE=XLSX
/FILE='C:\Users\Art\Desktop\excel3.xlsx'
/SHEET=name 'Sheet1'
/CELLRANGE=full
/READNAMES=on
/ASSUMEDSTRWIDTH=32767.
DATASET NAME DataSet3 WINDOW=FRONT.
list.
sort cases by id.
match file
file=dataset1
/file=dataset2
/file=dataset3
/by id
/map.
dataset name combo.
dataset activate combo.
list.

Art Kendall
Social Research Consultants

On 4/25/2010 3:32 PM, Mike Pritchard wrote:

Thanks

Mike

Mike Pritchard

Re: Creating missing variables without overwriting existing variables

In reply to this post by Maguin, Eugene

Hi Gene. Sorry I didn't see your message earlier.

Does the explanation in the reply to Art make sense to you? Essentially I
needed to add variables to allow me to use consistent syntax within Wincross
from the file manipulated in SPSS. The survey tool only exported variables
that contained data (this was a set of checkbox questions and there were
quite a few missing). I think the idea of merging from an SPSS file created
through Excel was correct, but somehow the SPSS file dropped the wrong
variables.

Mike

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
> Of Gene Maguin
> Sent: Sunday, April 25, 2010 2:14 PM
> To: [hidden email]
> Subject: Re: Creating missing variables without overwriting existing
> variables
>
> Mike,
>
> It might be helpful to all but certainly to me if you were to give some
> sort
> example to illustrate the problem. I'm sure some of the people on the
> list
> know exactly what you are describing, perhaps because they have to deal
> with
> it. For me, what you say doesn't quite make any sense because I've
> never
> seen data like you describe.
>
> Gene Maguin
>
> =====================
> 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

Maguin, Eugene

Re: Creating missing variables without overwriting existing variables

Mike,

In short (because it's late and I've been here all day): Ok, so you get a
file of numbers, how do you know which questions+subquestion/parts the
numbers go to. Your explanation of the file structure wasn't clear to me at
all. If Art or others understood, you don't need to explain it to me.

Gene Maguin

-----Original Message-----
From: Mike Pritchard [mailto:[hidden email]]
Sent: Sunday, April 25, 2010 9:01 PM
To: 'Gene Maguin'; [hidden email]
Subject: RE: Creating missing variables without overwriting existing
variables

Hi Gene. Sorry I didn't see your message earlier.

Does the explanation in the reply to Art make sense to you? Essentially I
needed to add variables to allow me to use consistent syntax within Wincross
from the file manipulated in SPSS. The survey tool only exported variables
that contained data (this was a set of checkbox questions and there were
quite a few missing). I think the idea of merging from an SPSS file created
through Excel was correct, but somehow the SPSS file dropped the wrong
variables.

Mike

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

Albert-Jan Roskam

Re: Creating missing variables without overwriting existing variables

In reply to this post by Mike Pritchard

Hi,

If you create a reference file that contains all the possible variables, e.g. reference.sav, you could do:

apply dictionary / from = 'reference.sav' / newvars.

Is this what you're looking for?

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 Sun, 4/25/10, Mike Pritchard <[hidden email]> wrote:

From: Mike Pritchard <[hidden email]>
Subject: Re: [SPSSX-L] Creating missing variables without overwriting existing variables
To: [hidden email]
Date: Sunday, April 25, 2010, 11:53 PM

Thanks Art. I’m not sure if the problem I encountered with the merge (later message) – other variables being deleted – was related, but what I meant by sparse data was that some variables were already there from the list of 52 for each option. What I was hoping to achieve was that new variables would be added, but that existing once would remain intact. I couldn’t figure out what was causing the deletion of variables that weren’t in the matched set, so I’m in the process of adding the variables by hand.

This particular set of questions caused me other problems with the export from the survey tool. The SPSS export is normally good, but this part was useless. It wasn’t a very complex question set, but perhaps complex enough to mess things up. The first question was a setup for a list of organizations to compare against. Then subsequent questions asked about particular attributes for the organizations checked in the first question, including Other/Specifies. The SPSS export only included results for option 1. The CSV export had complete results, but only for those checked. It probably would have been worth me adding the other variables into the CSV file before the original merge. Another option would have been to change each Wincross table to only have rows for populated variables, but that seemed more error prone.

I live and (sometimes) learn.

Mike

From: Art Kendall [mailto:[hidden email]]
Sent: Sunday, April 25, 2010 2:29 PM
To: Mike Pritchard
Cc: [hidden email]
Subject: Re: [SPSSX-L] Creating missing variables without overwriting existing variables

I am not sure I completely follow. See if the syntax below does what you want.

I created 3 excel files

id

q37_1

Q37_2

q37_3

1001

5

4

3

1002

2

1

3

1003

3

4

3

id

q38_1

Q38_2

q39_2

1001

7

12

25

1002

8

11

26

1003

9

13

27

1004

6

6

6

1005

3

3

1006

12

id

q37_1

q37_3

2001

22

51

2002

23

52

3003

24

53

GET DATA /TYPE=XLSX
/FILE='C:\Users\Art\Desktop\excel1.xlsx'
/SHEET=name 'Sheet1'
/CELLRANGE=full
/READNAMES=on
/ASSUMEDSTRWIDTH=32767.
DATASET NAME DataSet1 WINDOW=FRONT.
list.
sort cases by id.
GET DATA /TYPE=XLSX
/FILE='C:\Users\Art\Desktop\excel2.xlsx'
/SHEET=name 'Sheet1'
/CELLRANGE=full
/READNAMES=on
/ASSUMEDSTRWIDTH=32767.
DATASET NAME DataSet2 WINDOW=FRONT.
list.
sort cases by id.
GET DATA /TYPE=XLSX
/FILE='C:\Users\Art\Desktop\excel3.xlsx'
/SHEET=name 'Sheet1'
/CELLRANGE=full
/READNAMES=on
/ASSUMEDSTRWIDTH=32767.
DATASET NAME DataSet3 WINDOW=FRONT.
list.
sort cases by id.
match file
file=dataset1
/file=dataset2
/file=dataset3
/by id
/map.
dataset name combo.
dataset activate combo.
list.

Art Kendall
Social Research Consultants

On 4/25/2010 3:32 PM, Mike Pritchard wrote:
I’m working with some sparse data from an online survey, where the fields were only present for options that had responses. In addition to analysis with SPSS, I’m creating tables with Wincross. To make it easier/consistent, I’d like to have variables including all of the empty data.

I think there should be an easy way to do this within SPSS without manually creating a list. The variable names are of the form Q37_1_1, Q37_1_2, etc., Q37_2_1, Q37_2_2, etc., with each series having different missing names. The best idea I’ve had so far is to use Excel to create the list and then merge variables, but I wonder if there is an easier way using Syntax? The problem I see with COMPUTE is that it would erase data. I’m on SPSS V16

Thanks
Mike
_________________________________________________________________________
Mike Pritchard | mikep@... | 5 Circles Research | 425-444-3410
Helping companies delight their customers to increase profits