Multiple Responses to Categories
Locked
 
|
Hello List Sorry for this probably rather basic problem, which seems to be a common prerequisite to analyze multiple responses. I’ve got multiple responses (7), which gave me seven dichotomous variables funkt1 to funkt7. As I checked
the answers I figured out, that there have been chosen maximum three responses. Now I’d like to compute 3 new variables newvar1 to newvar3 inserting a value according to the number of multiple response (1 to 7) with a positive response (value 1). My aim is to compute at the end one categorical variable,
which contains the most frequent combinations of funkt1 to funkt7. For instance: funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 1 0 0 1 1 4 0 1 1 1 2 3 4 0 0 0 1 4 It looks like a loop and do if command? Thanks for help! Tom |
Re: Multiple Responses to Categories
|
|
Certainly a loop structure. Vector x=funkt1 to funkt7/y=newvar1 to newvar3. Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) y(#j)=#i. + if (x(#i) eq 1) #j=#j+1. End loop. What is this part about ‘…one categorical variable, which contains the most frequent combinations…’? Don’t you then want a variable like ‘combo’ with values of funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 combo 1 0 0 1 1 4 14 0 1 1 1 2 3 4 234 0 0 0 1 4 4 So would newvar1 to new3 are just be intermediates? If so, then why not this? Vector x=funkt1 to funkt7. String combo(a3). Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) substr(combo,#j,1)=1. + if (x(#i) eq 1) #j=#j+1. End loop. Gene Maguin From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Balmer, Thomas Hello List Sorry for this probably rather basic problem, which seems to be a common prerequisite to analyze multiple responses. I’ve got multiple responses (7), which gave me seven dichotomous variables funkt1 to funkt7. As I checked the answers I figured out, that there have been chosen maximum three responses. Now I’d like to compute 3 new variables newvar1 to newvar3 inserting a value according to the number of multiple response (1 to 7) with a positive response (value 1). My aim is to compute at the end one categorical variable, which contains the most frequent combinations of funkt1 to funkt7. For instance: funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 1 0 0 1 1 4 0 1 1 1 2 3 4 0 0 0 1 4 It looks like a loop and do if command? Thanks for help! Tom |
Re: Multiple Responses to Categories
|
In reply to this post by Tom
Tom There are two ways of dealing with this to start. 1: Use MULT RESPONSE in dichotomous mode. MULT RESPONSE groups difunkt (funkt1 to funkt7 (1)) /freq difunkt /tab difunkt by difunkt . and if you want the full three-way combinations add (in the last line) [ /difunkt by difunkt by difunkt. ] 2: Create seven new variables newfunkt1 to newfunkt7. DO REPEAT x = funkt1 to funkt7 /y = 1 to 7 /z = newfunkt1 to newfunkt7. RECODE x (1=y) into z. MULT RESPONSE groups mfunkt (newfunkt1 to newfunkt7 (1,7)) /tab mfunkt by mfunkt. [ /tab mfunkt by mfunkt by mfunkt. ] However (assuming all other values are zero and zero is not declared as missing) you can also create unique combinations of the seven binary codes COMPUTE combfunkt = funkt1*1000000+funkt2*100000+funkt3+10000+funkt4*1000 +funkt5*100+funkt6*10 + funkt7. FREQ combfunkt. By the same logic, you can produce a similar unique set of values using the recoded newfunkt~~~ COMPUTE combfunkt2 = newfunkt1*1000000+newfunkt2*100000+newfunkt3+10000+newfunkt4*1000 +newfunkt5*100+newfunkt6*10 + newfunkt7. FREQ combfunkt2. I have a set of tutorials on multiple response on page http://surveyresearch.weebly.com/33-multiple-response-mult-response.html on my website and there’s an example from real surveys in the 3rd slide show on page http://surveyresearch.weebly.com/old-dog-old-tricks.html. John F Hall (Mr) Email: [hidden email] Website: www.surveyresearch.weebly.com From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Balmer, Thomas Hello List Sorry for this probably rather basic problem, which seems to be a common prerequisite to analyze multiple responses. I’ve got multiple responses (7), which gave me seven dichotomous variables funkt1 to funkt7. As I checked the answers I figured out, that there have been chosen maximum three responses. Now I’d like to compute 3 new variables newvar1 to newvar3 inserting a value according to the number of multiple response (1 to 7) with a positive response (value 1). My aim is to compute at the end one categorical variable, which contains the most frequent combinations of funkt1 to funkt7. For instance: funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 1 0 0 1 1 4 0 1 1 1 2 3 4 0 0 0 1 4 It looks like a loop and do if command? Thanks for help! Tom |
Re: Multiple Responses to Categories
Administrator
|
In reply to this post by Tom
data list /x1 to x7 (7F1).
begin data 1000010 0111000 1100000 1100001 0100011 0000010 0011000 0100010 end data. COMPUTE ##=1. COMPUTE COMB=0. VECTOR X=X1 TO X7 / New(3). LOOP #=1 TO 7. + DO IF X(#)=1. + COMPUTE New(##)=#. + COMPUTE Comb=Comb*10+#. + COMPUTE ##=##+1. + END IF. END LOOP. EXE. FORMATS ALL (F3.0). LIST. X1 X2 X3 X4 X5 X6 X7 COMB NEW1 NEW2 NEW3 1 0 0 0 0 1 0 16 1 6 . 0 1 1 1 0 0 0 234 2 3 4 1 1 0 0 0 0 0 12 1 2 . 1 1 0 0 0 0 1 127 1 2 7 0 1 0 0 0 1 1 267 2 6 7 0 0 0 0 0 1 0 6 6 . . 0 0 1 1 0 0 0 34 3 4 . 0 1 0 0 0 1 0 26 2 6 . Number of cases read: 8 Number of cases listed: 8
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me. --- "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis." Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?" |
|
|
In reply to this post by Tom
Yes, Gene, your second solution is exactly what I’m heading for: to find all these combinations. But, trying your proposed syntax I got an error message (in German, sorry): GET FILE='I:\... DATASET NAME DatenSet1 WINDOW=FRONT. Vector x=FunktionRLP to FunktionSL. String combo(a3). Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) substr(combo,#j,1)=1. Fehler Nr. 4309 in Spalte 1024. Text: (Befehlsende) Ungültige Kombination von Datentypen in einer Zuordnung. Strings können nur alphanumerischen Variablen zugeordnet werden. Numerische und logische Werte können nur numerischen Variablen zugeordnet werden. Verwenden Sie die Funktion STRING oder NUMBER. Die Ausführung dieses Befehls wurde unterbrochen. + if (x(#i) eq 1) #j=#j+1. End loop. …”invalid combination of Types of Data in an assignment. Strings can be assigned only to alphanumeric variables (…)”. I do not understand this problem. The Variables Funktion are numeric.…which shouldn’t
be a problem? @ John Hall: I also tried your syntax to combinate the 1 and 0 from all the funkt-variables, but it doesn’t seem to work, yet, and, actually, I’d prefer the combination as Gene proposed: funkt1 funkt2 funkt3 funkt4 funkt5 funkt6 funkt7 comboJohn correct comboGene 1 1 0 0 0 0 0 1110000 1100000 12 0 0 0 0 0 1 0 10010 0000010 6 1 1 0 0 0 0 0 1110000 1100000 12 0 0 1 0 0 0 1 10002 0010001 37 1 1 0 0 0 0 0 1110000 1100000 12 Tom Von: SPSSX(r) Discussion
[hidden email]
Im Auftrag von Maguin, Eugene Certainly a loop structure. Vector x=funkt1 to funkt7/y=newvar1 to newvar3. Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) y(#j)=#i. + if (x(#i) eq 1) #j=#j+1. End loop. What is this part about ‘…one categorical variable, which contains the most frequent combinations…’? Don’t you then want a variable like ‘combo’ with values of
funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 combo 1 0 0 1 1 4 14 0 1 1 1 2 3 4 234 0 0 0 1 4 4 So would newvar1 to new3 are just be intermediates? If so, then why not this? Vector x=funkt1 to funkt7. String combo(a3). Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) substr(combo,#j,1)=1. + if (x(#i) eq 1) #j=#j+1. End loop. Gene Maguin From: SPSSX(r) Discussion
[hidden email]
On Behalf Of Balmer, Thomas Hello List Sorry for this probably rather basic problem, which seems to be a common prerequisite to analyze multiple responses. I’ve got multiple responses (7), which gave me seven dichotomous variables funkt1 to funkt7. As I checked
the answers I figured out, that there have been chosen maximum three responses. Now I’d like to compute 3 new variables newvar1 to newvar3 inserting a value according to the number of multiple response (1 to 7) with a positive response (value 1). My aim is to compute at the end one categorical variable,
which contains the most frequent combinations of funkt1 to funkt7. For instance: funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 1 0 0 1 1 4 0 1 1 1 2 3 4 0 0 0 1 4 It looks like a loop and do if command? Thanks for help! Tom |
Re: Multiple Responses to Categories
|
|
Thomas, My apologies. This line + if (x(#i) eq 1) substr(combo,#j,1)=1. Should be + if (x(#i) eq 1) substr(combo,#j,1)=’1’. Gene Maguin From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Balmer, Thomas Yes, Gene, your second solution is exactly what I’m heading for: to find all these combinations. But, trying your proposed syntax I got an error message (in German, sorry): GET FILE='I:\... DATASET NAME DatenSet1 WINDOW=FRONT. Vector x=FunktionRLP to FunktionSL. String combo(a3). Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) substr(combo,#j,1)=1. Fehler Nr. 4309 in Spalte 1024. Text: (Befehlsende) Ungültige Kombination von Datentypen in einer Zuordnung. Strings können nur alphanumerischen Variablen zugeordnet werden. Numerische und logische Werte können nur numerischen Variablen zugeordnet werden. Verwenden Sie die Funktion STRING oder NUMBER. Die Ausführung dieses Befehls wurde unterbrochen. + if (x(#i) eq 1) #j=#j+1. End loop. …”invalid combination of Types of Data in an assignment. Strings can be assigned only to alphanumeric variables (…)”. I do not understand this problem. The Variables Funktion are numeric.…which shouldn’t be a problem? @ John Hall: I also tried your syntax to combinate the 1 and 0 from all the funkt-variables, but it doesn’t seem to work, yet, and, actually, I’d prefer the combination as Gene proposed: funkt1 funkt2 funkt3 funkt4 funkt5 funkt6 funkt7 comboJohn correct comboGene 1 1 0 0 0 0 0 1110000 1100000 12 0 0 0 0 0 1 0 10010 0000010 6 1 1 0 0 0 0 0 1110000 1100000 12 0 0 1 0 0 0 1 10002 0010001 37 1 1 0 0 0 0 0 1110000 1100000 12 Tom Von: SPSSX(r) Discussion [hidden email] Im Auftrag von Maguin, Eugene Certainly a loop structure. Vector x=funkt1 to funkt7/y=newvar1 to newvar3. Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) y(#j)=#i. + if (x(#i) eq 1) #j=#j+1. End loop. What is this part about ‘…one categorical variable, which contains the most frequent combinations…’? Don’t you then want a variable like ‘combo’ with values of funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 combo 1 0 0 1 1 4 14 0 1 1 1 2 3 4 234 0 0 0 1 4 4 So would newvar1 to new3 are just be intermediates? If so, then why not this? Vector x=funkt1 to funkt7. String combo(a3). Compute #j=1. Loop #i=1 to 7. + if (x(#i) eq 1) substr(combo,#j,1)=1. + if (x(#i) eq 1) #j=#j+1. End loop. Gene Maguin From: SPSSX(r) Discussion [hidden email] On Behalf Of Balmer, Thomas Hello List Sorry for this probably rather basic problem, which seems to be a common prerequisite to analyze multiple responses. I’ve got multiple responses (7), which gave me seven dichotomous variables funkt1 to funkt7. As I checked the answers I figured out, that there have been chosen maximum three responses. Now I’d like to compute 3 new variables newvar1 to newvar3 inserting a value according to the number of multiple response (1 to 7) with a positive response (value 1). My aim is to compute at the end one categorical variable, which contains the most frequent combinations of funkt1 to funkt7. For instance: funkt1 funkt2 funkt3 funkt4 … newvar1 newvar2 newvar3 1 0 0 1 1 4 0 1 1 1 2 3 4 0 0 0 1 4 It looks like a loop and do if command? Thanks for help! Tom |
| Free forum by Nabble | Edit this page |