SPSSX Discussion

Re: mult response

classic Classic list List threaded Threaded
♦
♦
 2 messages Options
John F Hall
Reply | Threaded
Open this post in threaded view
♦
♦
|

Re: mult response

John F Hall
Muir
 
Copying this to list as it has general applications.
 
How many cases?  Do you have the raw data in machine readable format, or are you still at the coding stage?
 
If the latter, then you could do a grid a bit like the very first exercise in Block 1 on my site.
 
1.1.1  Questionnaire on pre-course interests and experience [23 Aug 2010: *.doc, 35 kb] 
         (Print up and complete, with up to 9 of your fellow students and/or colleagues)
 
1.2.1   Data transfer sheet  [24 Aug 2010: *.doc, 74 kb]
          (Print up, then write in the coded data from your completed questionnaire(s))
 
SPSS can handle column binary data, but it would be simpler and easier for you to allocate 7 variables and code them as 0,1 or 1-7.  If there are not too many cases you could type data direct into the SPSS data editor.
 
I've used your sample data here and tested it.  Only got one screenshot (MS Word won't let me copy the screenshots: keeps doing the last one it copied from SPSS) so I'll send you the full Word doc off list.  Anyway I've done a workaround as well.
 
 
Method 1 (Codes 1-7, any column)
 
Type data as appears into data editor regardless of codes or columns. i.e 7 codes, but only 4 columns
 
1 2 3 4
 
1 2 3
1 3 5 7
1
2 5 6 7
2 3
3 5
 
Syntax:
 
title 'Method1' .
mult resp groups
 hangout1 var00001 to var00004 (1,7)
 /freq hangout1 .
 

hangout1 Frequencies

 

Responses

Percent of Cases

N

Percent

hangout1a

1

3

18.8%

50.0%

2

3

18.8%

50.0%

3

4

25.0%

66.7%

5

3

18.8%

50.0%

6

1

6.3%

16.7%

7

2

12.5%

33.3%

Total

16

100.0%

266.7%

a. Group

 
Method 2
 
Type data into seven columns, using original codes .
 

Method 2 (codes 1-7, one column per code)

 

1 2 3 4 5 6 7

 

1 2 3
1   3   5   7
1
  2     5 6 7
  2 3
    3 4 5

 

Syntax:
 
title 'Method2' .
mult resp groups  hangout2 (var00001 to var00007 (1,7))
 /freq hangout2 .
 

hangout2 Frequencies

 

Responses

Percent of Cases

N

Percent

hangout2a

1

3

18.8%

50.0%

2

3

18.8%

50.0%

3

4

25.0%

66.7%

5

3

18.8%

50.0%

6

1

6.3%

16.7%

7

2

12.5%

33.3%

Total

16

100.0%

266.7%

a. Group

 
The two methods above require value labels only for the first variable in the list.  Note that SPSS does not report empty cells.
 
Method 3 (in dichotomous mode)
 
Type data into seven columns, using binary codes 0 or 1 .
 

<IMG src="file://C:\DOCUME~1\Owner\LOCALS~1\Temp\msohtml1\01\clip_image002.jpg" width=623 height=240 v:shapes="_x0000_i1025">

 
This method requires labels for each variable.
 
title 'Method3 - dichotomous mode' .
mult resp groups  hangout3 (var00001 to var00007 (1))
 /freq hangout3 .
 

hangout3 Frequencies

 

Responses

Percent of Cases

N

Percent

hangout3a

VAR00001

3

18.8%

50.0%

VAR00002

3

18.8%

50.0%

VAR00003

4

25.0%

66.7%

VAR00005

3

18.8%

50.0%

VAR00006

1

6.3%

16.7%

VAR00007

2

12.5%

33.3%

Total

16

100.0%

266.7%

a. Dichotomy group tabulated at value 1.

 
----- Original Message -----
Sent: Monday, October 04, 2010 12:30 PM
Subject: RE: mult response

Hi John,
had a quick look but it could be I need to restructure the variable before proceeding.
 
the question asked young people where they hung out with friends and 7 possible options were provided - they were asked to tick all that applied
 
 during the data input, these responses have been entered in one column, separated by ',' - so I could have:
1,2,3
1,3,5,7
1
2,5,6,7
2,3
3,5
 
and so on -
 
now my question is - do I need to recode into seven different variables or ....
 
thanks for any advice
 
best wishes
Muir
 
Dr Muir Houston
Lecturer
DACE
Faculty of Education
University of Glasgow
0141-330-4699

From: John F Hall
Sent: Tue 14/09/2010 17:24
To: Muir Houston
Subject: mult response

Muir
 
Here's the link to the slide show.  It's better than I remember it (all done in hell of a hurry with no previous experiance of PowerPoint).  There's a lot on MR.  In the text, for "binary" read "dichotomous" mode.
 
 
The the mult resp bits of the full presentation are on pp 27 -35 on:
 
 
John
John F Hall
Reply | Threaded
Open this post in threaded view
♦
♦
|

Re: mult response

John F Hall
Correction; syntax for method 1 should have read:
 
title 'Method1' .
mult resp groups
 hangout1 (var00001 to var00004 (1,7) )
 /freq hangout1 .
Free forum by Nabble Edit this page