Pairs of Correlations Macro
Locked
 
|
Hello List Members, I am attempting to write a macro to run several correlations. Specifically, I have two lists of variables, list1 and list2, and I want to get the correlations
between variables that occupy the same position in the two lists. For example, list1 is the variables X, Y, & Z, and list2 is the variables A, B, & C, and I only want to get these correlations (X,A), (Y,B), and, (Z,C).
I need this to work for a large number of variables (e.g., 20 in each list) when the variables are not consecutive in the file and are not subscripted. I have tried specifying a macro with three arguments: one for the number of variables pairs, one referring to the first list, and one for the second list, but
this does not work because I can't figure out how to use a loop to access the appropriate positions within each list.
Any suggestions? Thanks in advance, Jason _____________________________________________________________ Jason W. Beckstead, Ph.D. Professor/Quantitative Methodologist University of South Florida College of Public Health 13201 Bruce B. Downs Blvd., MDC56, Tampa, FL 33612, USA Statistical Editor, International Journal of Nursing Studies phone: +1.813.974.7667 fax: +1.813.974.4719 personal website: http://personal.health.usf.edu/jbeckste/ International Journal of Nursing Studies http://www.elsevier.com/ijns |
Re: Pairs of Correlations Macro
|
|
Hi, Jason, I did not test this (no SPSS at home) but see if it works: DEFINE !cor(list1 = !CHAREND("/") /list2 = !CMDEND) !LET !n1 = !NULL /*counter for list 1*/ !DO !var1 !IN (!list1) !LET !n1 = !CONCAT(!n1,1) !LET !n2 = !NULL /*counter for list 2*/ !DO !var2 !IN (!list2) !LET !n2 = !CONCAT(!n2,1) !IF !LENGTH(!n1) !EQ !LENGTH(!n2) !THEN CORRELATIONS /VARIABLES = !var1 !var2 /PRINT=TWOTAIL NOSIG /MISSING=PAIRWISE. !IFEND !DOEND !DOEND !ENDDEFINE. !cor list1 = a b c / list2 = d e f. Mario Giesel Munich, Germany
Am Mittwoch, 4. September 2019, 20:05:37 MESZ hat Beckstead, Jason <[hidden email]> Folgendes geschrieben:
Hello List Members, I am attempting to write a macro to run several correlations. Specifically, I have two lists of variables, list1 and list2, and I want to get the correlations between variables that occupy the same position in the two lists. For example, list1 is the variables X, Y, & Z, and list2 is the variables A, B, & C, and I only want to get these correlations (X,A), (Y,B), and, (Z,C). I need this to work for a large number of variables (e.g., 20 in each list) when the variables are not consecutive in the file and are not subscripted. I have tried specifying a macro with three arguments: one for the number of variables pairs, one referring to the first list, and one for the second list, but this does not work because I can't figure out how to use a loop to access the appropriate positions within each list.
Any suggestions?
Thanks in advance, Jason
_____________________________________________________________ Jason W. Beckstead, Ph.D. Professor/Quantitative Methodologist University of South Florida College of Public Health 13201 Bruce B. Downs Blvd., MDC56, Tampa, FL 33612, USA Statistical Editor, International Journal of Nursing Studies phone: +1.813.974.7667 fax: +1.813.974.4719 personal website: http://personal.health.usf.edu/jbeckste/ International Journal of Nursing Studies http://www.elsevier.com/ijns
|
Re: Pairs of Correlations Macro
|
|
In reply to this post by Beckstead, Jason
Here is a non clunky way to do it.
DEFINE !corDMM(!POS !CHAREND("/") /!POS= !CMDEND) !IF (!1 !NE !NULL) !THEN CORRELATIONS /VARIABLES = !HEAD(!1) !HEAD(!2). !corDMM !TAIL(!1)/ !TAIL( !2). !IFEND !ENDDEFINE. !corDMM a b c / d e f. Hello List Members, I am attempting to write a macro to run several correlations. Specifically, I have two lists of variables, list1 and list2, and I want to get the correlations between variables that occupy the same position in the two lists. For example, list1 is the variables X, Y, & Z, and list2 is the variables A, B, & C, and I only want to get these correlations (X,A), (Y,B), and, (Z,C). I need this to work for a large number of variables (e.g., 20 in each list) when the variables are not consecutive in the file and are not subscripted. I have tried specifying a macro with three arguments: one for the number of variables pairs, one referring to the first list, and one for the second list, but this does not work because I can't figure out how to use a loop to access the appropriate positions within each list. Any suggestions? Thanks in advance, Jason _____________________________________________________________ Jason W. Beckstead, Ph.D. Professor/Quantitative Methodologist University of South Florida College of Public Health 13201 Bruce B. Downs Blvd., MDC56, Tampa, FL 33612, USA Statistical Editor, International Journal of Nursing Studies phone: +1.813.974.7667 fax: +1.813.974.4719 personal website: http://personal.health.usf.edu/jbeckste/ International Journal of Nursing Studies http://www.elsevier.com/ijns ===================== 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 ===================== 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 |
|
|
Nice use of recursion, David. Next you could post a MATRIX solution. On Wed, Sep 4, 2019 at 1:29 PM David Marso <[hidden email]> wrote: Here is a non clunky way to do it. |
Administrator
|
Hear, hear! Very nice indeed!
Jon Peck wrote > Nice use of recursion, David. Next you could post a MATRIX solution. > > On Wed, Sep 4, 2019 at 1:29 PM David Marso < > Dmmugaspss@ > > wrote: > >> Here is a non clunky way to do it. >> >> DEFINE !corDMM(!POS !CHAREND("/") /!POS= !CMDEND) >> !IF (!1 !NE !NULL) !THEN >> CORRELATIONS >> /VARIABLES = !HEAD(!1) !HEAD(!2). >> !corDMM !TAIL(!1)/ !TAIL( !2). >> !IFEND >> !ENDDEFINE. >> >> !corDMM a b c / d e f. >> >> Hello List Members, >> >> I am attempting to write a macro to run several correlations. >> Specifically, I have two lists of variables, list1 and list2, and I want >> to >> get the correlations between variables that occupy the same position in >> the >> two lists. >> >> For example, list1 is the variables X, Y, & Z, and list2 is the variables >> A, B, & C, and I only want to get these correlations (X,A), (Y,B), and, >> (Z,C). >> >> I need this to work for a large number of variables (e.g., 20 in each >> list) when the variables are not consecutive in the file and are not >> subscripted. >> >> I have tried specifying a macro with three arguments: one for the number >> of variables pairs, one referring to the first list, and one for the >> second >> list, but this does not work because I can't figure out how to use a loop >> to access the appropriate positions within each list. >> >> >> >> Any suggestions? >> >> >> >> Thanks in advance, >> >> Jason >> >> >> >> _____________________________________________________________ >> >> Jason W. Beckstead, Ph.D. >> >> Professor/Quantitative Methodologist >> >> University of South Florida College of Public Health >> >> 13201 Bruce B. Downs Blvd., MDC56, Tampa, FL 33612, USA >> >> Statistical Editor, International Journal of Nursing Studies >> >> phone: +1.813.974.7667 fax: +1.813.974.4719 >> >> personal website: http://personal.health.usf.edu/jbeckste/ >> >> International Journal of Nursing Studies http://www.elsevier.com/ijns >> >> >> >> ===================== To manage your subscription to SPSSX-L, send a >> message to > LISTSERV@.UGA > (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 > LISTSERV@.UGA > (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 >> > LISTSERV@.UGA > (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 >> > > > -- > Jon K Peck > jkpeck@ > > ===================== > To manage your subscription to SPSSX-L, send a message to > LISTSERV@.UGA > (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 ----- -- Bruce Weaver [hidden email] http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." NOTE: My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
|
|
This process can be used to pair variables for any comparison. David posted a macro using recursion on Dec. 11, 2013, which computes both correlations and one-way anova.
DEFINE DoSomething (!POS !CMDEND).
CORRELATION !HEAD(!1) WITH !TAIL(!1) .
!ENDDEFINE .
DEFINE DoSomethingElse (!POS !CMDEND).
ONEWAY !HEAD(!1) BY !TAIL(!1) .
!ENDDEFINE .
DEFINE Parallel
(!POS !CHAREND ("/") / !POS !CHAREND ("/") / !POS !CMDEND ).
!UNQUOTE(!1) !HEAD(!2) !HEAD(!3) .
!IF (!TAIL(!2) !NE !NULL) !THEN
Parallel !1 / !TAIL(!2) / !TAIL(!3).
!IFEND
!ENDDEFINE.
DATA LIST FREE / a b c d e f.
BEGIN DATA
1 2 3 2 3 4
5 6 7 8 9 1
2 3 4 5 1 2
3 3 5 6 7 8
END DATA.
Parallel 'DoSomething' /a b c /d e f.
Parallel 'DoSomethingElse' /a b c /d e f.
For the correlations, you only need the first Define-!enddefine, the sample data, and the macro call line.
DEFINE DoSomething (!POS !CMDEND).
CORRELATION !HEAD(!1) WITH !TAIL(!1) .
!ENDDEFINE .
DATA LIST FREE / a b c d e f.
BEGIN DATA
1 2 3 2 3 4
5 6 7 8 9 1
2 3 4 5 1 2
3 3 5 6 7 8
END DATA.
Parallel 'DoSomething' /a b c /d e f.
Brian
From: SPSSX(r) Discussion <[hidden email]> on behalf of Bruce Weaver <[hidden email]>
Sent: Wednesday, September 4, 2019 4:01:07 PM To: [hidden email] <[hidden email]> Subject: Re: Pairs of Correlations Macro Hear, hear! Very nice indeed!
Jon Peck wrote > Nice use of recursion, David. Next you could post a MATRIX solution. > > On Wed, Sep 4, 2019 at 1:29 PM David Marso < > Dmmugaspss@ > > wrote: > >> Here is a non clunky way to do it. >> >> DEFINE !corDMM(!POS !CHAREND("/") /!POS= !CMDEND) >> !IF (!1 !NE !NULL) !THEN >> CORRELATIONS >> /VARIABLES = !HEAD(!1) !HEAD(!2). >> !corDMM !TAIL(!1)/ !TAIL( !2). >> !IFEND >> !ENDDEFINE. >> >> !corDMM a b c / d e f. >> >> Hello List Members, >> >> I am attempting to write a macro to run several correlations. >> Specifically, I have two lists of variables, list1 and list2, and I want >> to >> get the correlations between variables that occupy the same position in >> the >> two lists. >> >> For example, list1 is the variables X, Y, & Z, and list2 is the variables >> A, B, & C, and I only want to get these correlations (X,A), (Y,B), and, >> (Z,C). >> >> I need this to work for a large number of variables (e.g., 20 in each >> list) when the variables are not consecutive in the file and are not >> subscripted. >> >> I have tried specifying a macro with three arguments: one for the number >> of variables pairs, one referring to the first list, and one for the >> second >> list, but this does not work because I can't figure out how to use a loop >> to access the appropriate positions within each list. >> >> >> >> Any suggestions? >> >> >> >> Thanks in advance, >> >> Jason >> >> >> >> _____________________________________________________________ >> >> Jason W. Beckstead, Ph.D. >> >> Professor/Quantitative Methodologist >> >> University of South Florida College of Public Health >> >> 13201 Bruce B. Downs Blvd., MDC56, Tampa, FL 33612, USA >> >> Statistical Editor, International Journal of Nursing Studies >> >> phone: +1.813.974.7667 fax: +1.813.974.4719 >> >> personal website: http://personal.health.usf.edu/jbeckste/ >> >> International Journal of Nursing Studies http://www.elsevier.com/ijns >> >> >> >> ===================== To manage your subscription to SPSSX-L, send a >> message to > LISTSERV@.UGA > (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 > LISTSERV@.UGA > (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 >> > LISTSERV@.UGA > (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 >> > > > -- > Jon K Peck > jkpeck@ > > ===================== > To manage your subscription to SPSSX-L, send a message to > LISTSERV@.UGA > (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 ----- -- Bruce Weaver [hidden email] http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." NOTE: My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
In reply to this post by David Marso-2
I like David's idea of recursive macros. Without irony, Jon.
At least, it is creative. Often probably useful. 04.09.2019 22:29, David Marso пишет: > DEFINE !corDMM(!POS !CHAREND("/") /!POS= !CMDEND) > !IF (!1 !NE !NULL) !THEN > CORRELATIONS > /VARIABLES = !HEAD(!1) !HEAD(!2). > !corDMM !TAIL(!1)/ !TAIL( !2). > !IFEND > !ENDDEFINE. > > !corDMM a b c / d e f. ===================== 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 |
Re: Pairs of Correlations Macro
|
|
In reply to this post by Bruce Weaver
Small correction: A pair of parentheses were missing ... DEFINE !cor(list1 = !CHAREND("/") /list2 = !CMDEND) !LET !n1 = !NULL /*counter for list 1*/ !DO !var1 !IN (!list1) !LET !n1 = !CONCAT(!n1,1) !LET !n2 = !NULL /*counter for list 2*/ !DO !var2 !IN (!list2) !LET !n2 = !CONCAT(!n2,1) !IF (!LENGTH(!n1) !EQ !LENGTH(!n2)) !THEN CORRELATIONS /VARIABLES = !var1 !var2 /PRINT=TWOTAIL NOSIG /MISSING=PAIRWISE. !IFEND !DOEND !DOEND !ENDDEFINE. Mario Giesel Munich, Germany
Am Donnerstag, 5. September 2019, 02:10:58 MESZ hat Bruce Weaver <[hidden email]> Folgendes geschrieben:
Hear, hear! Very nice indeed! Jon Peck wrote > Nice use of recursion, David. Next you could post a MATRIX solution. > > On Wed, Sep 4, 2019 at 1:29 PM David Marso < > Dmmugaspss@ > > wrote: > >> Here is a non clunky way to do it. >> >> DEFINE !corDMM(!POS !CHAREND("/") /!POS= !CMDEND) >> !IF (!1 !NE !NULL) !THEN >> CORRELATIONS >> /VARIABLES = !HEAD(!1) !HEAD(!2). >> !corDMM !TAIL(!1)/ !TAIL( !2). >> !IFEND >> !ENDDEFINE. >> >> !corDMM a b c / d e f. >> >> Hello List Members, >> >> I am attempting to write a macro to run several correlations. >> Specifically, I have two lists of variables, list1 and list2, and I want >> to >> get the correlations between variables that occupy the same position in >> the >> two lists. >> >> For example, list1 is the variables X, Y, & Z, and list2 is the variables >> A, B, & C, and I only want to get these correlations (X,A), (Y,B), and, >> (Z,C). >> >> I need this to work for a large number of variables (e.g., 20 in each >> list) when the variables are not consecutive in the file and are not >> subscripted. >> >> I have tried specifying a macro with three arguments: one for the number >> of variables pairs, one referring to the first list, and one for the >> second >> list, but this does not work because I can't figure out how to use a loop >> to access the appropriate positions within each list. >> >> >> >> Any suggestions? >> >> >> >> Thanks in advance, >> >> Jason >> >> >> >> _____________________________________________________________ >> >> Jason W. Beckstead, Ph.D. >> >> Professor/Quantitative Methodologist >> >> University of South Florida College of Public Health >> >> 13201 Bruce B. Downs Blvd., MDC56, Tampa, FL 33612, USA >> >> Statistical Editor, International Journal of Nursing Studies >> >> phone: +1.813.974.7667 fax: +1.813.974.4719 >> >> personal website: http://personal.health.usf.edu/jbeckste/ >> >> International Journal of Nursing Studies http://www.elsevier.com/ijns >> >> >> >> ===================== 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 >> >> ===================== >> 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 >> > > > -- > Jon K Peck > jkpeck@ > > ===================== > 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 ----- -- Bruce Weaver [hidden email] http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." NOTE: My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
In reply to this post by Beckstead, Jason
My internet-esp suggests that the two lists have versions of
the same variables, and so you might be very happy with using
the t-test command, for PAIRS= .
And even the variables do not represent two versions of the
same variable, which the paired t-test is appropriate for, you
will get a moderately compact list of the correlations.
--
Rich Ulrich
From: SPSSX(r) Discussion <[hidden email]> on behalf of Beckstead, Jason <[hidden email]>
Sent: Wednesday, September 4, 2019 2:05 PM To: [hidden email] <[hidden email]> Subject: Pairs of Correlations Macro Hello List Members, I am attempting to write a macro to run several correlations. Specifically, I have two lists of variables, list1 and list2, and I want to get the correlations between variables that occupy the same position in the two lists. For example, list1 is the variables X, Y, & Z, and list2 is the variables A, B, & C, and I only want to get these correlations (X,A), (Y,B), and, (Z,C). I need this to work for a large number of variables (e.g., 20 in each list) when the variables are not consecutive in the file and are not subscripted. I have tried specifying a macro with three arguments: one for the number of variables pairs, one referring to the first list, and one for the second list, but this does not work because I can't figure out how to use a loop to access the appropriate positions within each list.
Any suggestions?
Thanks in advance, Jason
_____________________________________________________________ Jason W. Beckstead, Ph.D. Professor/Quantitative Methodologist University of South Florida College of Public Health 13201 Bruce B. Downs Blvd., MDC56, Tampa, FL 33612, USA Statistical Editor, International Journal of Nursing Studies phone: +1.813.974.7667 fax: +1.813.974.4719 personal website: http://personal.health.usf.edu/jbeckste/ International Journal of Nursing Studies http://www.elsevier.com/ijns
|
|
Administrator
|
Yes, well spotted, Rich. Here are some examples from the FM:
T-TEST PAIRS=TEACHER CONSTRUC MANAGER. T-TEST PAIRS=TEACHER MANAGER WITH CONSTRUC ENGINEER. T-TEST PAIRS=TEACHER MANAGER WITH CONSTRUC ENGINEER (PAIRED). * The first T-TEST compares TEACHER with CONSTRUC, TEACHER with MANAGER, and CONSTRUC with MANAGER. * The second T-TEST compares TEACHER with CONSTRUC, TEACHER with ENGINEER, MANAGER with CONSTRUC, and MANAGER with ENGINEER. TEACHER is not compared with MANAGER, and CONSTRUC is not compared with ENGINEER. * The third T-TEST compares TEACHER with CONSTRUC and MANAGER with ENGINEER. Source: https://www.ibm.com/support/knowledgecenter/en/SSLVMB_26.0.0/statistics_reference_project_ddita/spss/base/syn_t-test_pairs.html#syn_t-test_pairs One could also use OMS to suppress all of the unwanted t-test output. I don't have SPSS here, so can't test an example right now, but I think it would be something like this: OMS /IF COMMANDS = ['t-test'] /SELECT ALL EXCEPT = ['Paired Samples Correlations'] /DESTINATION VIEWER = NO /TAG = 'CorrOnly'. T-TEST PAIRS=X Y Z WITH A B C (PAIRED). OMSEND TAG = 'CorrOnly'. If no one else checks it first, I'll try this tomorrow when I'm back in the office. Once again, well spotted, Rich. Rich Ulrich wrote > My internet-esp suggests that the two lists have versions of > the same variables, and so you might be very happy with using > the t-test command, for PAIRS= . > > And even the variables do not represent two versions of the > same variable, which the paired t-test is appropriate for, you > will get a moderately compact list of the correlations. > > -- > Rich Ulrich ----- -- Bruce Weaver [hidden email] http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." NOTE: My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
|
Administrator
|
This post was updated on .
Here's a tested version that works.
DATA LIST LIST / X Y Z A B C (6F5.0). BEGIN DATA 2 5 4 7 3 9 8 3 5 4 6 7 1 2 3 4 5 6 6 5 4 3 2 1 END DATA. OMS /SELECT TABLES /IF COMMANDS=['T-Test'] SUBTYPES=['Paired Samples Statistics' 'Paired Samples Test'] /DESTINATION VIEWER=NO /TAG = 'CorrOnly'. T-TEST PAIRS=X Y Z WITH A B C (PAIRED). OMSEND TAG = 'CorrOnly'. OUTPUT (pivot table copied as plain text): Paired Samples Correlations N Correlation Sig. Pair 1 X & A 4 -.495 .505 Pair 2 Y & B 4 -.852 .148 Pair 3 Z & C 4 .120 .880 One added benefit of this approach is that the results are in a single table. Bruce Weaver wrote > Yes, well spotted, Rich. Here are some examples from the FM: > > T-TEST PAIRS=TEACHER CONSTRUC MANAGER. > T-TEST PAIRS=TEACHER MANAGER WITH CONSTRUC ENGINEER. > T-TEST PAIRS=TEACHER MANAGER WITH CONSTRUC ENGINEER (PAIRED). > > * The first T-TEST compares TEACHER with CONSTRUC, TEACHER with MANAGER, > and > CONSTRUC with MANAGER. > * The second T-TEST compares TEACHER with CONSTRUC, TEACHER with ENGINEER, > MANAGER with CONSTRUC, and MANAGER with ENGINEER. TEACHER is not compared > with MANAGER, and CONSTRUC is not compared with ENGINEER. > * The third T-TEST compares TEACHER with CONSTRUC and MANAGER with > ENGINEER. > > Source: > https://www.ibm.com/support/knowledgecenter/en/SSLVMB_26.0.0/statistics_reference_project_ddita/spss/base/syn_t-test_pairs.html#syn_t-test_pairs > > One could also use OMS to suppress all of the unwanted t-test output. I > don't have SPSS here, so can't test an example right now, but I think it > would be something like this: > > OMS /IF COMMANDS = ['t-test'] > /SELECT ALL EXCEPT = ['Paired Samples Correlations'] > /DESTINATION VIEWER = NO > /TAG = 'CorrOnly'. > > T-TEST PAIRS=X Y Z WITH A B C (PAIRED). > > OMSEND TAG = 'CorrOnly'. > > If no one else checks it first, I'll try this tomorrow when I'm back in > the > office. > > Once again, well spotted, Rich. > > > > Rich Ulrich wrote >> My internet-esp suggests that the two lists have versions of >> the same variables, and so you might be very happy with using >> the t-test command, for PAIRS= . >> >> And even the variables do not represent two versions of the >> same variable, which the paired t-test is appropriate for, you >> will get a moderately compact list of the correlations. >> >> -- >> Rich Ulrich > > > > > > ----- > -- > Bruce Weaver > bweaver@ > http://sites.google.com/a/lakeheadu.ca/bweaver/ > > "When all else fails, RTFM." > > NOTE: My Hotmail account is not monitored regularly. > To send me an e-mail, please use the address shown above. > > -- > Sent from: http://spssx-discussion.1045642.n5.nabble.com/ > > ===================== > To manage your subscription to SPSSX-L, send a message to > LISTSERV@.UGA > (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 ----- -- Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." NOTE: My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@LISTSERV.UGA.EDU (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
--
Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/ "When all else fails, RTFM." PLEASE NOTE THE FOLLOWING: 1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above. 2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/). |
Re: Pairs of Correlations Macro
|
|
In reply to this post by Beckstead, Jason
Still ultra klunky Mario.
===================== 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 |
«
Return to SPSSX Discussion
|
1 view|%1 views
| Free forum by Nabble | Edit this page |