Discussion:
multiple response set
(too old to reply)
F. Thomas
2015-01-06 23:32:26 UTC
Permalink
Hi,
I found the MRSETS command which allows to analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been implemented, according to
the manual.
So how to analyse mult response questions ? What can you do with MRSETS
when you have no Mult response frequencies or tables ?
Or do I miss something ?

And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used only by third party
software.
Could you please be more specific ? Which third party software do you mean ?

- ftr
Ben Pfaff
2015-01-08 05:54:10 UTC
Permalink
Post by F. Thomas
I found the MRSETS command which allows to analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been implemented, according to the
manual.
So how to analyse mult response questions ? What can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
Post by F. Thomas
And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used only by third party
software.
Could you please be more specific ? Which third party software do you mean ?
Software other than PSPP.
news
2015-01-08 13:36:57 UTC
Permalink
Post by Ben Pfaff
Post by F. Thomas
I found the MRSETS command which allows to analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been implemented, according to the
manual.
So how to analyse mult response questions ? What can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in survey
research and few stats programs have a proc to analyse them.

Having this opportunity in PSPP would strongly increase its usefulness
for a wider audience.
Post by Ben Pfaff
Post by F. Thomas
And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used only by third party
software.
Could you please be more specific ? Which third party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
Matthias Faeth
2015-01-08 14:22:00 UTC
Permalink
I would support that. Multi Response is the one procedure that lets me
stick to SPSS. I'm not a progammer but would help with testing and
comparing.

Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by news
Post by Ben Pfaff
Post by F. Thomas
I found the MRSETS command which allows to analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been implemented, according to the
manual.
So how to analyse mult response questions ? What can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in survey
research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its usefulness for
a wider audience.
And what does the cryptic sentence mean (manual p.113)
Post by Ben Pfaff
Post by F. Thomas
Otherwise, multiple response sets are currently used only by third party
software.
Could you please be more specific ? Which third party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
Alan Mead
2015-01-08 16:21:34 UTC
Permalink
I've used SPSS to analyze multiple response data for years (decades,
actually) but never used MULT RESPONSE. I was curious what I was
missing, so I watched this video:
and I'm still a bit
confused. You get the same data by running frequencies on the four
variables independently, right?

If each response is optional, then one thing that is a bit of a PITA is
detecting non-response, but that's not a big deal. For example, if the
four possible responses to Q12 are encoded 1/0 in Q12A, Q12B, Q12C, and
Q12D, then you can do this:

count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.

Everyone with Q12MISS=0 didn't respond to the question. For some
questions, this is more important than individual responses (other times
not).

I'm not arguing against including it in PSPP, I'm just curious why it's
an issue because it seems like it's really, really easy to get along
without. What am I missing?

BTW, there is another issue of multiple responses that DOESN'T work this
way. When you have a test question labeled "Mark all that apply" and if
your scoring is all or nothing then it's actually easier to handle this
as a string. If they marked A, B and E on Q12, you encode their
response as 'ABE'. Later you score it: "recode Q12 ('ABC'=1) (else=0)
into Q12.Scored." If you're going to give partial credit for individual
responses, it's usually easier to enter the individual responses as
independent variables, but you could create them using string
functions. So, again, SPSS without MULT RESPONSE seems perfectly
adequate and MULT RESPONSE doesn't actually handle all
multiple-responses situations.

-Alan
Post by Matthias Faeth
I would support that. Multi Response is the one procedure that lets me
stick to SPSS. I'm not a progammer but would help with testing and
comparing.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
I found the MRSETS command which allows to analyse
multiple reponse
questions;
But the MULT RESPONSE command has not yet been
implemented, according to the
manual.
So how to analyse mult response questions ? What can you
do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in
survey research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its
usefulness for a wider audience.
And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used only
by third party
software.
Could you please be more specific ? Which third party
software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.

science + technology = better workers

+815.588.3846 (Office)
+267.334.4143 (Mobile)

http://www.alanmead.org

Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
ftr
2015-01-08 16:54:22 UTC
Permalink
In fact the Multiple response procedure is particular useful because
stats programs are based on the statistical independence of observations
whereas in survey research you often have multiple response sets when
the same respondent has more than one answer to a question, i.e. the
cases are statistically dependent.

This is why I vote for the implementation of the mult response proc ,
for practical reasons and to increase the attractiveness of PSPP for
larger audiences .
Did you already work with multiple response questions ?

Take the example of drinks and age.
Usually you have one answer for the question: what do you drink ?

But in reality you drink Coke as well as water, beer, but not soda.

So the same person has several answers for the same question.

Now differentiate that by 3 age groups:

The table in MULT RESPONSE can show how many times the same cases
(=persons) in the low, the intermediate, the high age group drink drink
Coke AS WELL AS the another drink beer, either in percentage of cases (%
of persons in the low age group drink Coke, % beer, etc.) or how many
beer drinks are in the this age group.

Better than the Youtube video the text
<http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/3.3.2a1__spss_15___first_exercise_in_multiple_response.pdf>of
John Hall (see p.7) can provide you an idea.
http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/3.3.2a1__spss_15___first_exercise_in_multiple_response.pdf

- ftr
Post by Alan Mead
I've used SPSS to analyze multiple response data for years (decades,
actually) but never used MULT RESPONSE. I was curious what I was
http://youtu.be/-toBCDscCwQ and I'm still a bit
confused. You get the same data by running frequencies on the four
variables independently, right?
If each response is optional, then one thing that is a bit of a PITA
is detecting non-response, but that's not a big deal. For example, if
the four possible responses to Q12 are encoded 1/0 in Q12A, Q12B,
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question. For some
questions, this is more important than individual responses (other
times not).
I'm not arguing against including it in PSPP, I'm just curious why
it's an issue because it seems like it's really, really easy to get
along without. What am I missing?
BTW, there is another issue of multiple responses that DOESN'T work
this way. When you have a test question labeled "Mark all that apply"
and if your scoring is all or nothing then it's actually easier to
handle this as a string. If they marked A, B and E on Q12, you encode
their response as 'ABE'. Later you score it: "recode Q12 ('ABC'=1)
(else=0) into Q12.Scored." If you're going to give partial credit for
individual responses, it's usually easier to enter the individual
responses as independent variables, but you could create them using
string functions. So, again, SPSS without MULT RESPONSE seems
perfectly adequate and MULT RESPONSE doesn't actually handle all
multiple-responses situations.
-Alan
Post by Matthias Faeth
I would support that. Multi Response is the one procedure that lets
me stick to SPSS. I'm not a progammer but would help with testing and
comparing.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
I found the MRSETS command which allows to analyse
multiple reponse
questions;
But the MULT RESPONSE command has not yet been
implemented, according to the
manual.
So how to analyse mult response questions ? What can you
do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but
multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in
survey research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its
usefulness for a wider audience.
And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used only
by third party
software.
Could you please be more specific ? Which third party
software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing:http://www.iacat.org/jcat
Alan Mead
2015-01-08 17:25:33 UTC
Permalink
So, you're saying that the crosstabs capability is critical. I'm sure
you have a point, I don't use crosstabs much. I'll have to give that
some thought as to whether there is an easy workaround for that.

-Alan
Post by ftr
In fact the Multiple response procedure is particular useful because
stats programs are based on the statistical independence of
observations whereas in survey research you often have multiple
response sets when the same respondent has more than one answer to a
question, i.e. the cases are statistically dependent.
This is why I vote for the implementation of the mult response proc ,
for practical reasons and to increase the attractiveness of PSPP for
larger audiences .
Did you already work with multiple response questions ?
Take the example of drinks and age.
Usually you have one answer for the question: what do you drink ?
But in reality you drink Coke as well as water, beer, but not soda.
So the same person has several answers for the same question.
The table in MULT RESPONSE can show how many times the same cases
(=persons) in the low, the intermediate, the high age group drink
drink Coke AS WELL AS the another drink beer, either in percentage of
cases (% of persons in the low age group drink Coke, % beer, etc.) or
how many beer drinks are in the this age group.
Better than the Youtube video the text
<http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/3.3.2a1__spss_15___first_exercise_in_multiple_response.pdf>of
John Hall (see p.7) can provide you an idea.
http://surveyresearch.weebly.com/uploads/2/9/9/8/2998485/3.3.2a1__spss_15___first_exercise_in_multiple_response.pdf
- ftr
Post by Alan Mead
I've used SPSS to analyze multiple response data for years (decades,
actually) but never used MULT RESPONSE. I was curious what I was
http://youtu.be/-toBCDscCwQ and I'm still a bit
confused. You get the same data by running frequencies on the four
variables independently, right?
If each response is optional, then one thing that is a bit of a PITA
is detecting non-response, but that's not a big deal. For example,
if the four possible responses to Q12 are encoded 1/0 in Q12A, Q12B,
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question. For some
questions, this is more important than individual responses (other
times not).
I'm not arguing against including it in PSPP, I'm just curious why
it's an issue because it seems like it's really, really easy to get
along without. What am I missing?
BTW, there is another issue of multiple responses that DOESN'T work
this way. When you have a test question labeled "Mark all that apply"
and if your scoring is all or nothing then it's actually easier to
handle this as a string. If they marked A, B and E on Q12, you
encode their response as 'ABE'. Later you score it: "recode Q12
('ABC'=1) (else=0) into Q12.Scored." If you're going to give partial
credit for individual responses, it's usually easier to enter the
individual responses as independent variables, but you could create
them using string functions. So, again, SPSS without MULT RESPONSE
seems perfectly adequate and MULT RESPONSE doesn't actually handle
all multiple-responses situations.
-Alan
Post by Matthias Faeth
I would support that. Multi Response is the one procedure that lets
me stick to SPSS. I'm not a progammer but would help with testing
and comparing.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
I found the MRSETS command which allows to analyse
multiple reponse
questions;
But the MULT RESPONSE command has not yet been
implemented, according to the
manual.
So how to analyse mult response questions ? What can you
do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in
survey research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its
usefulness for a wider audience.
And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used
only by third party
software.
Could you please be more specific ? Which third party
software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.

science + technology = better workers

+815.588.3846 (Office)
+267.334.4143 (Mobile)

http://www.alanmead.org

Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
Alan Mead
2015-01-08 18:16:52 UTC
Permalink
FWIW, I don't understand your example and I tried to run your example,
but my license of SPSS does not include CTABLES.

If you wanted to know how many 1's in variables 83 to 84, you could use
these two lines:

count NUMLIKES = var82 to var84 (1).
FREQ/ NUMLIKES.

And I suspect that you could do a crosstabs with NUMLIKES and get the
same kind of information as CTABLES.

-Alan
Well I use MRSETS usually for open questions. Here the issue is, that
each code can be on any variable in the set in arbitrary order.
e.g. "Likes" get 3 possible variables var82 var83 var84.
Case A: 1 2 3
Case B: 4 1 5
Case C: 5 6 1
MRSETS
/mcgroup name=$Likes VARIABLES =var82 var83 var84.
And make a table which would tell me that 1 is in every case (for each
pack which is here var80)
CTABLES
/VLABELS VARIABLES=$likes DISPLAY=none
/table $likes by var80
/CATEGORIES VARIABLES=$likes totals=yes EMPTY=EXCLUDE
/TITLES TITLE = 'Likes Pack'
.
As far as I know, PSPP does not support this.
Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
I've used SPSS to analyze multiple response data for years
(decades, actually) but never used MULT RESPONSE. I was curious
http://youtu.be/-toBCDscCwQ and I'm still a bit
confused. You get the same data by running frequencies on the
four variables independently, right?
If each response is optional, then one thing that is a bit of a
PITA is detecting non-response, but that's not a big deal. For
example, if the four possible responses to Q12 are encoded 1/0 in
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question. For some
questions, this is more important than individual responses (other
times not).
I'm not arguing against including it in PSPP, I'm just curious why
it's an issue because it seems like it's really, really easy to
get along without. What am I missing?
BTW, there is another issue of multiple responses that DOESN'T
work this way. When you have a test question labeled "Mark all
that apply" and if your scoring is all or nothing then it's
actually easier to handle this as a string. If they marked A, B
and E on Q12, you encode their response as 'ABE'. Later you score
it: "recode Q12 ('ABC'=1) (else=0) into Q12.Scored." If you're
going to give partial credit for individual responses, it's
usually easier to enter the individual responses as independent
variables, but you could create them using string functions. So,
again, SPSS without MULT RESPONSE seems perfectly adequate and
MULT RESPONSE doesn't actually handle all multiple-responses
situations.
-Alan
Post by Matthias Faeth
I would support that. Multi Response is the one procedure that
lets me stick to SPSS. I'm not a progammer but would help with
testing and comparing.
Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
I found the MRSETS command which allows to analyse
multiple reponse
questions;
But the MULT RESPONSE command has not yet been
implemented, according to the
manual.
So how to analyse mult response questions ? What can
you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is
implemented to allow the
.sav file format to be more completely supported, but
multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used
in survey research and few stats programs have a proc to
analyse them.
Having this opportunity in PSPP would strongly increase its
usefulness for a wider audience.
And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used
only by third party
software.
Could you please be more specific ? Which third party
software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.

science + technology = better workers

+815.588.3846 (Office)
+267.334.4143 (Mobile)

http://www.alanmead.org

Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
Matthias Faeth
2015-01-08 20:44:37 UTC
Permalink
Yes I could use count, but the problem with open questions is that I
usually have a codeplan with up to 100 codes. That makes it tedious to use
your solution. And it would not produce 1 comprehensive table.



Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by Alan Mead
FWIW, I don't understand your example and I tried to run your example,
but my license of SPSS does not include CTABLES.
If you wanted to know how many 1's in variables 83 to 84, you could use
count NUMLIKES = var82 to var84 (1).
FREQ/ NUMLIKES.
And I suspect that you could do a crosstabs with NUMLIKES and get the same
kind of information as CTABLES.
-Alan
Well I use MRSETS usually for open questions. Here the issue is, that
each code can be on any variable in the set in arbitrary order.
e.g. "Likes" get 3 possible variables var82 var83 var84.
Case A: 1 2 3
Case B: 4 1 5
Case C: 5 6 1
MRSETS
/mcgroup name=$Likes VARIABLES =var82 var83 var84.
And make a table which would tell me that 1 is in every case (for each
pack which is here var80)
CTABLES
/VLABELS VARIABLES=$likes DISPLAY=none
/table $likes by var80
/CATEGORIES VARIABLES=$likes totals=yes EMPTY=EXCLUDE
/TITLES TITLE = 'Likes Pack'
.
As far as I know, PSPP does not support this.
Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by Alan Mead
I've used SPSS to analyze multiple response data for years (decades,
actually) but never used MULT RESPONSE. I was curious what I was missing,
so I watched this video: http://youtu.be/-toBCDscCwQ and
I'm still a bit confused. You get the same data by running frequencies on
the four variables independently, right?
If each response is optional, then one thing that is a bit of a PITA is
detecting non-response, but that's not a big deal. For example, if the
four possible responses to Q12 are encoded 1/0 in Q12A, Q12B, Q12C, and
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question. For some
questions, this is more important than individual responses (other times
not).
I'm not arguing against including it in PSPP, I'm just curious why it's
an issue because it seems like it's really, really easy to get along
without. What am I missing?
BTW, there is another issue of multiple responses that DOESN'T work this
way. When you have a test question labeled "Mark all that apply" and if
your scoring is all or nothing then it's actually easier to handle this as
a string. If they marked A, B and E on Q12, you encode their response as
'ABE'. Later you score it: "recode Q12 ('ABC'=1) (else=0) into
Q12.Scored." If you're going to give partial credit for individual
responses, it's usually easier to enter the individual responses as
independent variables, but you could create them using string functions.
So, again, SPSS without MULT RESPONSE seems perfectly adequate and MULT
RESPONSE doesn't actually handle all multiple-responses situations.
-Alan
I would support that. Multi Response is the one procedure that lets me
stick to SPSS. I'm not a progammer but would help with testing and
comparing.
Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by news
Post by Ben Pfaff
Post by F. Thomas
I found the MRSETS command which allows to analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been implemented, according to the
manual.
So how to analyse mult response questions ? What can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in survey
research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its usefulness
for a wider audience.
And what does the cryptic sentence mean (manual p.113)
Post by Ben Pfaff
Post by F. Thomas
Otherwise, multiple response sets are currently used only by third party
software.
Could you please be more specific ? Which third party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
Frans Houweling
2015-01-09 20:21:09 UTC
Permalink
Hi Matthias,
I use good ol' (SPSS 14) TABLES, so I don't usually save MRSETS but
define them on the fly with /MRGROUP; I guess CTABLES give similar results.
Often I prefer to translate my MR variables to MD format like this
(suppose codes with max code 100, and suppose max 10 answers):

VECTOR bin (100) /* max code */.
DO REPEAT mr = mr1 TO mr10 /* input vars */.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin100 /* max code again */ (SYSMIS=0) (1=1) /* or
for readability (1=100) */.

The tedious part is converting VAL LAB 1 "Alfa Romeo" 2 "Audi" etc. into
VAR LAB bin1 "Alfa Romeo" /bin2 "Audi" etc., but once this is done you
can use MEANS or DESCRIPTIVES in combination with SPLIT FILE to live
without (C)TABLES and without SPSS(C).
HTH
frans
Post by Matthias Faeth
Yes I could use count, but the problem with open questions is that I
usually have a codeplan with up to 100 codes. That makes it tedious to
use your solution. And it would not produce 1 comprehensive table.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
FWIW, I don't understand your example and I tried to run your
example, but my license of SPSS does not include CTABLES.
If you wanted to know how many 1's in variables 83 to 84, you
count NUMLIKES = var82 to var84 (1).
FREQ/ NUMLIKES.
And I suspect that you could do a crosstabs with NUMLIKES and get
the same kind of information as CTABLES.
-Alan
Well I use MRSETS usually for open questions. Here the issue is,
that each code can be on any variable in the set in arbitrary order.
e.g. "Likes" get 3 possible variables var82 var83 var84.
Case A: 1 2 3
Case B: 4 1 5
Case C: 5 6 1
MRSETS
/mcgroup name=$Likes VARIABLES =var82 var83 var84.
And make a table which would tell me that 1 is in every case (for
each pack which is here var80)
CTABLES
/VLABELS VARIABLES=$likes DISPLAY=none
/table $likes by var80
/CATEGORIES VARIABLES=$likes totals=yes EMPTY=EXCLUDE
/TITLES TITLE = 'Likes Pack'
.
As far as I know, PSPP does not support this.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
I've used SPSS to analyze multiple response data for years
(decades, actually) but never used MULT RESPONSE. I was
http://youtu.be/-toBCDscCwQ and I'm still a
bit confused. You get the same data by running frequencies
on the four variables independently, right?
If each response is optional, then one thing that is a bit of
a PITA is detecting non-response, but that's not a big deal.
For example, if the four possible responses to Q12 are
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question. For
some questions, this is more important than individual
responses (other times not).
I'm not arguing against including it in PSPP, I'm just
curious why it's an issue because it seems like it's really,
really easy to get along without. What am I missing?
BTW, there is another issue of multiple responses that
DOESN'T work this way. When you have a test question labeled
"Mark all that apply" and if your scoring is all or nothing
then it's actually easier to handle this as a string. If
they marked A, B and E on Q12, you encode their response as
'ABE'. Later you score it: "recode Q12 ('ABC'=1) (else=0)
into Q12.Scored." If you're going to give partial credit for
individual responses, it's usually easier to enter the
individual responses as independent variables, but you could
create them using string functions. So, again, SPSS without
MULT RESPONSE seems perfectly adequate and MULT RESPONSE
doesn't actually handle all multiple-responses situations.
-Alan
Post by Matthias Faeth
I would support that. Multi Response is the one procedure
that lets me stick to SPSS. I'm not a progammer but would
help with testing and comparing.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
On Wed, Jan 07, 2015 at 12:32:26AM +0100, F. Thomas
I found the MRSETS command which allows to
analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been
implemented, according to the
manual.
So how to analyse mult response questions ? What
can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is
implemented to allow the
.sav file format to be more completely supported,
but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely
used in survey research and few stats programs have a
proc to analyse them.
Having this opportunity in PSPP would strongly increase
its usefulness for a wider audience.
And what does the cryptic sentence mean (manual
p.113)
Otherwise, multiple response sets are currently
used only by third party
software.
Could you please be more specific ? Which third
party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing:http://www.iacat.org/jcat
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing:http://www.iacat.org/jcat
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
F. Thomas
2015-01-10 12:29:39 UTC
Permalink
This sounds interesting. Can you please translate this into a practical
example because I don't understand all of your moves ?

Cheers,
ftr
Post by Frans Houweling
Hi Matthias,
I use good ol' (SPSS 14) TABLES, so I don't usually save MRSETS but
define them on the fly with /MRGROUP; I guess CTABLES give similar results.
Often I prefer to translate my MR variables to MD format like this
VECTOR bin (100) /* max code */.
DO REPEAT mr = mr1 TO mr10 /* input vars */.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin100 /* max code again */ (SYSMIS=0) (1=1) /* or
for readability (1=100) */.
The tedious part is converting VAL LAB 1 "Alfa Romeo" 2 "Audi" etc.
into VAR LAB bin1 "Alfa Romeo" /bin2 "Audi" etc., but once this is
done you can use MEANS or DESCRIPTIVES in combination with SPLIT FILE
to live without (C)TABLES and without SPSS(C).
HTH
frans
Post by Matthias Faeth
Yes I could use count, but the problem with open questions is that I
usually have a codeplan with up to 100 codes. That makes it tedious
to use your solution. And it would not produce 1 comprehensive table.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
FWIW, I don't understand your example and I tried to run your
example, but my license of SPSS does not include CTABLES.
If you wanted to know how many 1's in variables 83 to 84, you
count NUMLIKES = var82 to var84 (1).
FREQ/ NUMLIKES.
And I suspect that you could do a crosstabs with NUMLIKES and get
the same kind of information as CTABLES.
-Alan
Well I use MRSETS usually for open questions. Here the issue is,
that each code can be on any variable in the set in arbitrary order.
e.g. "Likes" get 3 possible variables var82 var83 var84.
Case A: 1 2 3
Case B: 4 1 5
Case C: 5 6 1
MRSETS
/mcgroup name=$Likes VARIABLES =var82 var83 var84.
And make a table which would tell me that 1 is in every case
(for each pack which is here var80)
CTABLES
/VLABELS VARIABLES=$likes DISPLAY=none
/table $likes by var80
/CATEGORIES VARIABLES=$likes totals=yes EMPTY=EXCLUDE
/TITLES TITLE = 'Likes Pack'
.
As far as I know, PSPP does not support this.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
I've used SPSS to analyze multiple response data for years
(decades, actually) but never used MULT RESPONSE. I was
http://youtu.be/-toBCDscCwQ and I'm still a
bit confused. You get the same data by running frequencies
on the four variables independently, right?
If each response is optional, then one thing that is a bit
of a PITA is detecting non-response, but that's not a big
deal. For example, if the four possible responses to Q12 are
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question. For
some questions, this is more important than individual
responses (other times not).
I'm not arguing against including it in PSPP, I'm just
curious why it's an issue because it seems like it's really,
really easy to get along without. What am I missing?
BTW, there is another issue of multiple responses that
DOESN'T work this way. When you have a test question labeled
"Mark all that apply" and if your scoring is all or nothing
then it's actually easier to handle this as a string. If
they marked A, B and E on Q12, you encode their response as
'ABE'. Later you score it: "recode Q12 ('ABC'=1) (else=0)
into Q12.Scored." If you're going to give partial credit
for individual responses, it's usually easier to enter the
individual responses as independent variables, but you could
create them using string functions. So, again, SPSS without
MULT RESPONSE seems perfectly adequate and MULT RESPONSE
doesn't actually handle all multiple-responses situations.
-Alan
Post by Matthias Faeth
I would support that. Multi Response is the one procedure
that lets me stick to SPSS. I'm not a progammer but would
help with testing and comparing.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
On Wed, Jan 07, 2015 at 12:32:26AM +0100, F. Thomas
I found the MRSETS command which allows to
analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been
implemented, according to the
manual.
So how to analyse mult response questions ?
What can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is
implemented to allow the
.sav file format to be more completely supported,
but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a
widely used in survey research and few stats programs
have a proc to analyse them.
Having this opportunity in PSPP would strongly increase
its usefulness for a wider audience.
And what does the cryptic sentence mean (manual
p.113)
Otherwise, multiple response sets are currently
used only by third party
software.
Could you please be more specific ? Which third
party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing:http://www.iacat.org/jcat
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing:http://www.iacat.org/jcat
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
Frans Houweling
2015-01-10 17:48:42 UTC
Permalink
Hi ftr,
sorry if this syntax is long but I did not have any suitable sav file
at hand.


******* Create fake data N=200
************************************************.
DATA LIST FREE /dummy (F3.0).
BEGIN DATA
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6
7 8 9 0 1 2 3 4 5 6 7 8 9 0
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6
7 8 9 0 1 2 3 4 5 6 7 8 9 0
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6
7 8 9 0 1 2 3 4 5 6 7 8 9 0
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6
7 8 9 0 1 2 3 4 5 6 7 8 9 0
END DATA.

**** Make 10 variables with random data r1 TO r10.
VECTOR r (10).
LOOP #i = 1 TO 10.
+ COMPUTE r(#i) = RV.UNIFORM(0, 100).
END LOOP.
EXE.

**** Make 10 dichotomic vars m1 TO m10.
DO REPEAT r = r1 TO r10 / p = 27 67 80 34 11 60 21 44 8 59 / m = m1 TO m10.
+ DO IF MOD(RND(r), 5) NE 0.
+ COMPUTE m = r LE p.
+ END IF.
END REPEAT.
EXE.
VAR LAB m1 "Sports" /m2 "Food" /m3 "Cinema" /m4 "Books" /m5 "Art" /m6
"Music" /m7 "Nature"
/m8 "TV" /m9 "Politics" /m10 "Automobiles".

*** Make fake breakdown var.
COMPUTE segment = dummy GT 3.
VAR LAB segment "Segment".
VAL LAB segment 0 "Hi" 1 "Lo".

******* End create fake data
**************************************************.

DESCRIPTIVES m1 TO m10.

**** Table 1. SPSS only.
*TABLES
/MDGROUP = hobbies "" m1 TO m10 (1)
/PTOTAL = pt "Total"
/TABLE = hobbies BY pt + segment
/STATISTICS CPCT(hobbies (PCT5.1) "": segment)
/TITLE="Don't do this at home with PSPP".

*** Convert dichotomic MD to numeric MR.
*** I create 10 vars here but usually the MR answers are less than the
MD codes.
COMPUTE #k = 0.
VECTOR mr (10).
DO REPEAT md = m1 TO m10.
+ COMPUTE #k = #k + 1.
+ DO IF md.
+ COMPUTE #done = 0.
+ LOOP #i = 1 TO 10.
+ DO IF SYSMIS(mr(#i)) AND NOT #done.
+ COMPUTE mr(#i) = #k.
+ COMPUTE #done = 1.
+ END IF.
+ END LOOP.
+ END IF.
END REPEAT.
EXECUTE.
VAL LAB mr1 TO mr10 1 "Sports" 2 "Food" 3 "Cinema" 4 "Books" 5 "Art" 6
"Music" 7 "Nature"
8 "TV" 9 "Politics" 10 "Automobiles".

**** Table 2 should be same as Table 1. SPSS only.
*TABLES
/MRGROUP = hobbies "" mr1 TO mr10
/PTOTAL = pt "Total"
/TABLE = hobbies BY pt + segment
/STATISTICS CPCT(hobbies (PCT5.1) "": segment).

*** Convert numeric MR to dichotomic MD.
VECTOR bin (10) /* highest code in MR vars */.
DO REPEAT mr = mr1 TO mr10.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin10 (SYSMIS=0) (1=100).
VAR LAB bin1 "Sports" /bin2 "Food" /bin3 "Cinema" /bin4 "Books" /bin5
"Art" /bin6 "Music" /bin7 "Nature"
/bin8 "TV" /bin9 "Politics" /bin10 "Automobiles".

**** Same result again.
*TABLES
/MDGROUP = hobbies "" bin1 TO bin10 (100)
/PTOTAL = pt "Total"
/TABLE = hobbies BY pt + segment
/STATISTICS CPCT(hobbies (PCT5.1) "": segment).

*** PSPP alternatives.
** Would be nice if PSPP honoured FORMATS.
FORMATS bin1 TO bin10 (PCT5.1) /* does nothing ?*/.

MEANS bin1 TO bin10 BY segment.

SORT CASES BY segment.
SPLIT FILE BY segment.
DESCRIPTIVES bin1 TO bin10.
SPLIT FILE OFF.

** My favourite. Finish off in spreadsheet.
AGGREGATE OUTFILE = * /BREAK = segment
/bin1 TO bin10 = MEAN(bin1 TO bin10).
FLIP.
SAVE TRANSLATE /OUTFILE = "table.txt"
/TYPE=TAB /FIELDNAMES /REPLACE.

Regards
frans
Post by F. Thomas
This sounds interesting. Can you please translate this into a
practical example because I don't understand all of your moves ?
Cheers,
ftr
Post by Frans Houweling
Hi Matthias,
I use good ol' (SPSS 14) TABLES, so I don't usually save MRSETS but
define them on the fly with /MRGROUP; I guess CTABLES give similar results.
Often I prefer to translate my MR variables to MD format like this
VECTOR bin (100) /* max code */.
DO REPEAT mr = mr1 TO mr10 /* input vars */.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin100 /* max code again */ (SYSMIS=0) (1=1) /* or
for readability (1=100) */.
The tedious part is converting VAL LAB 1 "Alfa Romeo" 2 "Audi" etc.
into VAR LAB bin1 "Alfa Romeo" /bin2 "Audi" etc., but once this is
done you can use MEANS or DESCRIPTIVES in combination with SPLIT FILE
to live without (C)TABLES and without SPSS(C).
HTH
frans
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
Matthias Faeth
2015-01-10 21:09:41 UTC
Permalink
seems likes magic to me :->


Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by Frans Houweling
Hi ftr,
sorry if this syntax is long but I did not have any suitable sav file at
hand.
******* Create fake data N=200
************************************************.
DATA LIST FREE /dummy (F3.0).
BEGIN DATA
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7
8 9 0 1 2 3 4 5 6 7 8 9 0
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7
8 9 0 1 2 3 4 5 6 7 8 9 0
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7
8 9 0 1 2 3 4 5 6 7 8 9 0
1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7
8 9 0 1 2 3 4 5 6 7 8 9 0
END DATA.
**** Make 10 variables with random data r1 TO r10.
VECTOR r (10).
LOOP #i = 1 TO 10.
+ COMPUTE r(#i) = RV.UNIFORM(0, 100).
END LOOP.
EXE.
**** Make 10 dichotomic vars m1 TO m10.
DO REPEAT r = r1 TO r10 / p = 27 67 80 34 11 60 21 44 8 59 / m = m1 TO m10.
+ DO IF MOD(RND(r), 5) NE 0.
+ COMPUTE m = r LE p.
+ END IF.
END REPEAT.
EXE.
VAR LAB m1 "Sports" /m2 "Food" /m3 "Cinema" /m4 "Books" /m5 "Art" /m6
"Music" /m7 "Nature"
/m8 "TV" /m9 "Politics" /m10 "Automobiles".
*** Make fake breakdown var.
COMPUTE segment = dummy GT 3.
VAR LAB segment "Segment".
VAL LAB segment 0 "Hi" 1 "Lo".
******* End create fake data
**************************************************.
DESCRIPTIVES m1 TO m10.
**** Table 1. SPSS only.
*TABLES
/MDGROUP = hobbies "" m1 TO m10 (1)
/PTOTAL = pt "Total"
/TABLE = hobbies BY pt + segment
/STATISTICS CPCT(hobbies (PCT5.1) "": segment)
/TITLE="Don't do this at home with PSPP".
*** Convert dichotomic MD to numeric MR.
*** I create 10 vars here but usually the MR answers are less than the MD
codes.
COMPUTE #k = 0.
VECTOR mr (10).
DO REPEAT md = m1 TO m10.
+ COMPUTE #k = #k + 1.
+ DO IF md.
+ COMPUTE #done = 0.
+ LOOP #i = 1 TO 10.
+ DO IF SYSMIS(mr(#i)) AND NOT #done.
+ COMPUTE mr(#i) = #k.
+ COMPUTE #done = 1.
+ END IF.
+ END LOOP.
+ END IF.
END REPEAT.
EXECUTE.
VAL LAB mr1 TO mr10 1 "Sports" 2 "Food" 3 "Cinema" 4 "Books" 5 "Art" 6
"Music" 7 "Nature"
8 "TV" 9 "Politics" 10 "Automobiles".
**** Table 2 should be same as Table 1. SPSS only.
*TABLES
/MRGROUP = hobbies "" mr1 TO mr10
/PTOTAL = pt "Total"
/TABLE = hobbies BY pt + segment
/STATISTICS CPCT(hobbies (PCT5.1) "": segment).
*** Convert numeric MR to dichotomic MD.
VECTOR bin (10) /* highest code in MR vars */.
DO REPEAT mr = mr1 TO mr10.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin10 (SYSMIS=0) (1=100).
VAR LAB bin1 "Sports" /bin2 "Food" /bin3 "Cinema" /bin4 "Books" /bin5
"Art" /bin6 "Music" /bin7 "Nature"
/bin8 "TV" /bin9 "Politics" /bin10 "Automobiles".
**** Same result again.
*TABLES
/MDGROUP = hobbies "" bin1 TO bin10 (100)
/PTOTAL = pt "Total"
/TABLE = hobbies BY pt + segment
/STATISTICS CPCT(hobbies (PCT5.1) "": segment).
*** PSPP alternatives.
** Would be nice if PSPP honoured FORMATS.
FORMATS bin1 TO bin10 (PCT5.1) /* does nothing ?*/.
MEANS bin1 TO bin10 BY segment.
SORT CASES BY segment.
SPLIT FILE BY segment.
DESCRIPTIVES bin1 TO bin10.
SPLIT FILE OFF.
** My favourite. Finish off in spreadsheet.
AGGREGATE OUTFILE = * /BREAK = segment
/bin1 TO bin10 = MEAN(bin1 TO bin10).
FLIP.
SAVE TRANSLATE /OUTFILE = "table.txt"
/TYPE=TAB /FIELDNAMES /REPLACE.
Regards
frans
This sounds interesting. Can you please translate this into a practical
example because I don't understand all of your moves ?
Cheers,
ftr
Hi Matthias,
I use good ol' (SPSS 14) TABLES, so I don't usually save MRSETS but
define them on the fly with /MRGROUP; I guess CTABLES give similar results.
Often I prefer to translate my MR variables to MD format like this
VECTOR bin (100) /* max code */.
DO REPEAT mr = mr1 TO mr10 /* input vars */.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin100 /* max code again */ (SYSMIS=0) (1=1) /* or for
readability (1=100) */.
The tedious part is converting VAL LAB 1 "Alfa Romeo" 2 "Audi" etc. into
VAR LAB bin1 "Alfa Romeo" /bin2 "Audi" etc., but once this is done you can
use MEANS or DESCRIPTIVES in combination with SPLIT FILE to live without
(C)TABLES and without SPSS(C).
HTH
frans
_______________________________________________
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
Matthias Faeth
2015-01-10 13:10:25 UTC
Permalink
Hi Frans

great piece of code, that makes life easier.

I was not aware that PSPP supports advanced features like vector, which I
have heard of bit never used before.

Thanks
Matthias

Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by Frans Houweling
Hi Matthias,
I use good ol' (SPSS 14) TABLES, so I don't usually save MRSETS but
define them on the fly with /MRGROUP; I guess CTABLES give similar results.
Often I prefer to translate my MR variables to MD format like this
VECTOR bin (100) /* max code */.
DO REPEAT mr = mr1 TO mr10 /* input vars */.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin100 /* max code again */ (SYSMIS=0) (1=1) /* or for
readability (1=100) */.
The tedious part is converting VAL LAB 1 "Alfa Romeo" 2 "Audi" etc. into
VAR LAB bin1 "Alfa Romeo" /bin2 "Audi" etc., but once this is done you can
use MEANS or DESCRIPTIVES in combination with SPLIT FILE to live without
(C)TABLES and without SPSS(C).
HTH
frans
Yes I could use count, but the problem with open questions is that I
usually have a codeplan with up to 100 codes. That makes it tedious to use
your solution. And it would not produce 1 comprehensive table.
Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by Alan Mead
FWIW, I don't understand your example and I tried to run your example,
but my license of SPSS does not include CTABLES.
If you wanted to know how many 1's in variables 83 to 84, you could use
count NUMLIKES = var82 to var84 (1).
FREQ/ NUMLIKES.
And I suspect that you could do a crosstabs with NUMLIKES and get the
same kind of information as CTABLES.
-Alan
Well I use MRSETS usually for open questions. Here the issue is, that
each code can be on any variable in the set in arbitrary order.
e.g. "Likes" get 3 possible variables var82 var83 var84.
Case A: 1 2 3
Case B: 4 1 5
Case C: 5 6 1
MRSETS
/mcgroup name=$Likes VARIABLES =var82 var83 var84.
And make a table which would tell me that 1 is in every case (for each
pack which is here var80)
CTABLES
/VLABELS VARIABLES=$likes DISPLAY=none
/table $likes by var80
/CATEGORIES VARIABLES=$likes totals=yes EMPTY=EXCLUDE
/TITLES TITLE = 'Likes Pack'
.
As far as I know, PSPP does not support this.
Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by Alan Mead
I've used SPSS to analyze multiple response data for years (decades,
actually) but never used MULT RESPONSE. I was curious what I was missing,
so I watched this video: http://youtu.be/-toBCDscCwQ
and I'm still a bit confused. You get the same data by running frequencies
on the four variables independently, right?
If each response is optional, then one thing that is a bit of a PITA is
detecting non-response, but that's not a big deal. For example, if the
four possible responses to Q12 are encoded 1/0 in Q12A, Q12B, Q12C, and
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question. For some
questions, this is more important than individual responses (other times
not).
I'm not arguing against including it in PSPP, I'm just curious why it's
an issue because it seems like it's really, really easy to get along
without. What am I missing?
BTW, there is another issue of multiple responses that DOESN'T work this
way. When you have a test question labeled "Mark all that apply" and if
your scoring is all or nothing then it's actually easier to handle this as
a string. If they marked A, B and E on Q12, you encode their response as
'ABE'. Later you score it: "recode Q12 ('ABC'=1) (else=0) into
Q12.Scored." If you're going to give partial credit for individual
responses, it's usually easier to enter the individual responses as
independent variables, but you could create them using string functions.
So, again, SPSS without MULT RESPONSE seems perfectly adequate and MULT
RESPONSE doesn't actually handle all multiple-responses situations.
-Alan
I would support that. Multi Response is the one procedure that lets me
stick to SPSS. I'm not a progammer but would help with testing and
comparing.
Matthias FÀth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Post by news
Post by Ben Pfaff
Post by F. Thomas
I found the MRSETS command which allows to analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been implemented, according to the
manual.
So how to analyse mult response questions ? What can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in
survey research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its usefulness
for a wider audience.
And what does the cryptic sentence mean (manual p.113)
Post by Ben Pfaff
Post by F. Thomas
Otherwise, multiple response sets are currently used only by third party
software.
Could you please be more specific ? Which third party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing: http://www.iacat.org/jcat
_______________________________________________
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
F. Thomas
2015-01-10 13:51:04 UTC
Permalink
The vector variable is described in the manual on p.109.
Post by Matthias Faeth
Hi Frans
great piece of code, that makes life easier.
I was not aware that PSPP supports advanced features like vector,
which I have heard of bit never used before.
Thanks
Matthias
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973
m: 0171-9832175
Hi Matthias,
I use good ol' (SPSS 14) TABLES, so I don't usually save MRSETS
but define them on the fly with /MRGROUP; I guess CTABLES give
similar results.
Often I prefer to translate my MR variables to MD format like this
VECTOR bin (100) /* max code */.
DO REPEAT mr = mr1 TO mr10 /* input vars */.
+ DO IF NOT MISSING(mr).
+ COMPUTE bin(mr) = 1.
+ END IF.
END REPEAT.
RECODE bin1 TO bin100 /* max code again */ (SYSMIS=0) (1=1) /*
or for readability (1=100) */.
The tedious part is converting VAL LAB 1 "Alfa Romeo" 2 "Audi"
etc. into VAR LAB bin1 "Alfa Romeo" /bin2 "Audi" etc., but once
this is done you can use MEANS or DESCRIPTIVES in combination with
SPLIT FILE to live without (C)TABLES and without SPSS(C).
HTH
frans
Post by Matthias Faeth
Yes I could use count, but the problem with open questions is
that I usually have a codeplan with up to 100 codes. That makes
it tedious to use your solution. And it would not produce 1
comprehensive table.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
FWIW, I don't understand your example and I tried to run your
example, but my license of SPSS does not include CTABLES.
If you wanted to know how many 1's in variables 83 to 84, you
count NUMLIKES = var82 to var84 (1).
FREQ/ NUMLIKES.
And I suspect that you could do a crosstabs with NUMLIKES and
get the same kind of information as CTABLES.
-Alan
Well I use MRSETS usually for open questions. Here the issue
is, that each code can be on any variable in the set in
arbitrary order.
e.g. "Likes" get 3 possible variables var82 var83 var84.
Case A: 1 2 3
Case B: 4 1 5
Case C: 5 6 1
MRSETS
/mcgroup name=$Likes VARIABLES =var82 var83 var84.
And make a table which would tell me that 1 is in every case
(for each pack which is here var80)
CTABLES
/VLABELS VARIABLES=$likes DISPLAY=none
/table $likes by var80
/CATEGORIES VARIABLES=$likes totals=yes EMPTY=EXCLUDE
/TITLES TITLE = 'Likes Pack'
.
As far as I know, PSPP does not support this.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
I've used SPSS to analyze multiple response data for
years (decades, actually) but never used MULT RESPONSE.
I was curious what I was missing, so I watched this
video: http://youtu.be/-toBCDscCwQ and
I'm still a bit confused. You get the same data by
running frequencies on the four variables independently,
right?
If each response is optional, then one thing that is a
bit of a PITA is detecting non-response, but that's not
a big deal. For example, if the four possible responses
to Q12 are encoded 1/0 in Q12A, Q12B, Q12C, and Q12D,
count Q12MISS = Q12A A12B Q12C Q12D (1).
execute.
Everyone with Q12MISS=0 didn't respond to the question.
For some questions, this is more important than
individual responses (other times not).
I'm not arguing against including it in PSPP, I'm just
curious why it's an issue because it seems like it's
really, really easy to get along without. What am I
missing?
BTW, there is another issue of multiple responses that
DOESN'T work this way. When you have a test question
labeled "Mark all that apply" and if your scoring is all
or nothing then it's actually easier to handle this as a
string. If they marked A, B and E on Q12, you encode
their response as 'ABE'. Later you score it: "recode
Q12 ('ABC'=1) (else=0) into Q12.Scored." If you're
going to give partial credit for individual responses,
it's usually easier to enter the individual responses as
independent variables, but you could create them using
string functions. So, again, SPSS without MULT RESPONSE
seems perfectly adequate and MULT RESPONSE doesn't
actually handle all multiple-responses situations.
-Alan
Post by Matthias Faeth
I would support that. Multi Response is the one
procedure that lets me stick to SPSS. I'm not a
progammer but would help with testing and comparing.
Matthias Fäth
Im Mediapark 12
50670 Köln
t: 0221-2907973 <tel:0221-2907973>
m: 0171-9832175 <tel:0171-9832175>
On Wed, Jan 07, 2015 at 12:32:26AM +0100, F.
I found the MRSETS command which allows to
analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet
been implemented, according to the
manual.
So how to analyse mult response questions ?
What can you do with MRSETS when
you have no Mult response frequencies or
tables ?
There is no such functionality yet. MRSETS is
implemented to allow the
.sav file format to be more completely
supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a
widely used in survey research and few stats
programs have a proc to analyse them.
Having this opportunity in PSPP would strongly
increase its usefulness for a wider audience.
And what does the cryptic sentence mean
(manual p.113)
Otherwise, multiple response sets are
currently used only by third party
software.
Could you please be more specific ? Which
third party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing:http://www.iacat.org/jcat
--
Alan D. Mead, Ph.D.
President, Talent Algorithms Inc.
science + technology = better workers
+815.588.3846 (Office)
+267.334.4143 (Mobile)
http://www.alanmead.org
Announcing the Journal of Computerized Adaptive Testing (JCAT), a
peer-reviewed electronic journal designed to advance the science and
practice of computerized adaptive testing:http://www.iacat.org/jcat
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
_______________________________________________
Pspp-users mailing list
https://lists.gnu.org/mailman/listinfo/pspp-users
Ben Pfaff
2015-01-08 15:55:58 UTC
Permalink
Post by news
Post by Ben Pfaff
Post by F. Thomas
I found the MRSETS command which allows to analyse multiple reponse
questions;
But the MULT RESPONSE command has not yet been implemented, according to the
manual.
So how to analyse mult response questions ? What can you do with MRSETS when
you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow the
.sav file format to be more completely supported, but multiple response
sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in survey
research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its usefulness for a
wider audience.
It's good to know that. No one has mentioned it before, to the best of
my recollection.
Post by news
Post by Ben Pfaff
Post by F. Thomas
And what does the cryptic sentence mean (manual p.113)
Otherwise, multiple response sets are currently used only by third party
software.
Could you please be more specific ? Which third party software do you mean ?
Software other than PSPP.
This was already evident to me. But which one ? SPSS ?
That's all I know about.
Marshall DeBerry
2015-01-08 16:23:37 UTC
Permalink
I would just echo that after SPSS introduced the MULT RESPONSE syntax in the
late 70's-early 80's (?), it was of immense help to my work on the DOJ/BJS
National Crime Victimization Survey in the production of reports/analyses.
It would be very helpful if this command syntax/feature could be added to
PSPP in the near future.
Thanks,
Marshall

-----Original Message-----
From: pspp-users-bounces+marsdeb=***@gnu.org
[mailto:pspp-users-bounces+marsdeb=***@gnu.org] On Behalf Of Ben
Pfaff
Sent: Thursday, January 08, 2015 10:56 AM
To: news
Cc: pspp-***@gnu.org
Subject: Re: multiple response set
Post by news
Post by Ben Pfaff
Post by F. Thomas
I found the MRSETS command which allows to analyse multiple reponse
questions; But the MULT RESPONSE command has not yet been
implemented, according to the manual.
So how to analyse mult response questions ? What can you do with
MRSETS when you have no Mult response frequencies or tables ?
There is no such functionality yet. MRSETS is implemented to allow
the .sav file format to be more completely supported, but multiple
response sets are not otherwise useful.
This is a pity. The multiple response format is a widely used in
survey research and few stats programs have a proc to analyse them.
Having this opportunity in PSPP would strongly increase its usefulness
for a wider audience.
It's good to know that. No one has mentioned it before, to the best of my
recollection.
Continue reading on narkive:
Loading...