**FILES FOR BOTH DATASETS ATTACHED (RQ1 is the old file I am comparing the new file to)**

Hello all, I received the following error upon running my syntax: "There are at least two records with the same values for the subject and within-subject variables."

The reason this baffles me is because I ran a similar syntax before with no problem, albeit I transformed this data into a team-level analysis, and so I no longer have an individual participant measure. I am trying to predict the likelihood for teams to select ideas that fit into three groups (categories that represent how much the idea diverges from the current paradigm of solving that problem). Below, please see my syntax from my PREVIOUS analysis at the individual level (just for main effects):

GENLIN Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeanNumber ParticipantNumber
(ORDER=ASCENDING) WITH IndividualKAI IndividualPsychSafety
/MODEL DesignTask Semester Instructor IndividualKAI IndividualPsychSafety
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=ParticipantNumber*TeanNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

Now for the NEW syntax manipulated to perform a similar analysis at the TEAM level. Note that we no longer have IndividualKAI, as that has been transformed into KAIDiversity and KAIElevation (standard dev. and mean measures for the team). We also have psychological safety as a team-level measure as PS_TP3:

GENLIN Final_Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeamNumber
(ORDER=ASCENDING) WITH KAIElevation KAIDiversity PS_TP3
/MODEL DesignTask Semester Instructor KAIElevation KAIDiversity PS_TP3
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

---

I've been trying to figure this out for a while. I thought maybe something was off with "REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count," but I tried to match the dataset file to what I had for my individual level analysis (the old one that worked). Any suggestions would be greatly appreciated!

------------------------------
Courtney Cole
------------------------------

#SPSSStatistics

Hi, Courtney. I tried this:

sort cases by TeamNumber Semester count.
do if lag(TeamNumber)=TeamNumber and lag(Semester)=Semester and lag(count)=count.
print / $casenum.
end if.
execute.​

and SPSS returned these case numbers as duplicates:

29, 33 35, 37, 40, 44. 47, 51, 53, 56, 61, 74, 80, 83, 90,
94, 99, 103, 112, 115, 118, 120, 122, 124, 127, 131, 139,
140, 144,  146, and 151.

So perhaps that's the problem....?

------------------------------
Rick Marcantonio
Quality Assurance
IBM
------------------------------

Original Message:
Sent: Thu February 10, 2022 01:57 PM
From: Courtney Cole
Subject: Syntax issue? Same values for subject and within-subject variables?

**FILES FOR BOTH DATASETS ATTACHED (RQ1 is the old file I am comparing the new file to)**

Hello all, I received the following error upon running my syntax: "There are at least two records with the same values for the subject and within-subject variables."

The reason this baffles me is because I ran a similar syntax before with no problem, albeit I transformed this data into a team-level analysis, and so I no longer have an individual participant measure. I am trying to predict the likelihood for teams to select ideas that fit into three groups (categories that represent how much the idea diverges from the current paradigm of solving that problem). Below, please see my syntax from my PREVIOUS analysis at the individual level (just for main effects):

GENLIN Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeanNumber ParticipantNumber
(ORDER=ASCENDING) WITH IndividualKAI IndividualPsychSafety
/MODEL DesignTask Semester Instructor IndividualKAI IndividualPsychSafety
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=ParticipantNumber*TeanNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

Now for the NEW syntax manipulated to perform a similar analysis at the TEAM level. Note that we no longer have IndividualKAI, as that has been transformed into KAIDiversity and KAIElevation (standard dev. and mean measures for the team). We also have psychological safety as a team-level measure as PS_TP3:

GENLIN Final_Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeamNumber
(ORDER=ASCENDING) WITH KAIElevation KAIDiversity PS_TP3
/MODEL DesignTask Semester Instructor KAIElevation KAIDiversity PS_TP3
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

---

I've been trying to figure this out for a while. I thought maybe something was off with "REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count," but I tried to match the dataset file to what I had for my individual level analysis (the old one that worked). Any suggestions would be greatly appreciated!

------------------------------
Courtney Cole
------------------------------
#SPSSStatistics

Hello Rick!
Thank you very much for your response. I tried running the syntax you posted for myself and removed Semester, so there is definitely some issue between TeamNumber and count. Is there a particular way I should have my within-subjects factor set up, or would you say GEE is not a viable option for my analysis? I know I can look at the selection process from the perspective of percentages of ideas selected by teams per each of the three categories (this would result in three analyses as a hierarchical linear regression), but if I wanted to look at my data more from the perspective of GEE or a multinomial logistic regression, I feel like my dataset as is would probably not work. I am okay with switching to hierarchical linear regression, but I just want to make sure I exhaust all my other options for the right reasons.

Thank you!

Courtney

------------------------------
Courtney Cole
------------------------------

Original Message:
Sent: Fri February 11, 2022 11:39 AM
From: Rick Marcantonio
Subject: Syntax issue? Same values for subject and within-subject variables?

Hi, Courtney. I tried this:

sort cases by TeamNumber Semester count.do if lag(TeamNumber)=TeamNumber and lag(Semester)=Semester and lag(count)=count.print / $casenum.end if.execute.​

and SPSS returned these case numbers as duplicates:

29, 33 35, 37, 40, 44. 47, 51, 53, 56, 61, 74, 80, 83, 90,
94, 99, 103, 112, 115, 118, 120, 122, 124, 127, 131, 139,
140, 144,  146, and 151.

So perhaps that's the problem....?

------------------------------
Rick Marcantonio
Quality Assurance
IBM

Original Message:
Sent: Thu February 10, 2022 01:57 PM
From: Courtney Cole
Subject: Syntax issue? Same values for subject and within-subject variables?

**FILES FOR BOTH DATASETS ATTACHED (RQ1 is the old file I am comparing the new file to)**

Hello all, I received the following error upon running my syntax: "There are at least two records with the same values for the subject and within-subject variables."

The reason this baffles me is because I ran a similar syntax before with no problem, albeit I transformed this data into a team-level analysis, and so I no longer have an individual participant measure. I am trying to predict the likelihood for teams to select ideas that fit into three groups (categories that represent how much the idea diverges from the current paradigm of solving that problem). Below, please see my syntax from my PREVIOUS analysis at the individual level (just for main effects):

GENLIN Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeanNumber ParticipantNumber
(ORDER=ASCENDING) WITH IndividualKAI IndividualPsychSafety
/MODEL DesignTask Semester Instructor IndividualKAI IndividualPsychSafety
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=ParticipantNumber*TeanNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

Now for the NEW syntax manipulated to perform a similar analysis at the TEAM level. Note that we no longer have IndividualKAI, as that has been transformed into KAIDiversity and KAIElevation (standard dev. and mean measures for the team). We also have psychological safety as a team-level measure as PS_TP3:

GENLIN Final_Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeamNumber
(ORDER=ASCENDING) WITH KAIElevation KAIDiversity PS_TP3
/MODEL DesignTask Semester Instructor KAIElevation KAIDiversity PS_TP3
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

---

I've been trying to figure this out for a while. I thought maybe something was off with "REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count," but I tried to match the dataset file to what I had for my individual level analysis (the old one that worked). Any suggestions would be greatly appreciated!

------------------------------
Courtney Cole
------------------------------
#SPSSStatistics

From what little I know, if the behavior you are studying naturally has a within-ss component to it, then GEE would be the way to go. However, I'm no GEE expert, Courtney, so I can't say with any authority. Sounds like a question either for your advisor or someone on your committee who is familiar with your research and the precise hypothes(i|e)s you want to answer.

------------------------------
Rick Marcantonio
Quality Assurance
IBM
------------------------------

Original Message:
Sent: Fri February 11, 2022 02:06 PM
From: Courtney Cole
Subject: Syntax issue? Same values for subject and within-subject variables?

Hello Rick!
Thank you very much for your response. I tried running the syntax you posted for myself and removed Semester, so there is definitely some issue between TeamNumber and count. Is there a particular way I should have my within-subjects factor set up, or would you say GEE is not a viable option for my analysis? I know I can look at the selection process from the perspective of percentages of ideas selected by teams per each of the three categories (this would result in three analyses as a hierarchical linear regression), but if I wanted to look at my data more from the perspective of GEE or a multinomial logistic regression, I feel like my dataset as is would probably not work. I am okay with switching to hierarchical linear regression, but I just want to make sure I exhaust all my other options for the right reasons.

Thank you!

Courtney

------------------------------
Courtney Cole

Original Message:
Sent: Fri February 11, 2022 11:39 AM
From: Rick Marcantonio
Subject: Syntax issue? Same values for subject and within-subject variables?

Hi, Courtney. I tried this:

sort cases by TeamNumber Semester count.do if lag(TeamNumber)=TeamNumber and lag(Semester)=Semester and lag(count)=count.print / $casenum.end if.execute.​

and SPSS returned these case numbers as duplicates:

29, 33 35, 37, 40, 44. 47, 51, 53, 56, 61, 74, 80, 83, 90,
94, 99, 103, 112, 115, 118, 120, 122, 124, 127, 131, 139,
140, 144,  146, and 151.

So perhaps that's the problem....?

------------------------------
Rick Marcantonio
Quality Assurance
IBM

Original Message:
Sent: Thu February 10, 2022 01:57 PM
From: Courtney Cole
Subject: Syntax issue? Same values for subject and within-subject variables?

**FILES FOR BOTH DATASETS ATTACHED (RQ1 is the old file I am comparing the new file to)**

Hello all, I received the following error upon running my syntax: "There are at least two records with the same values for the subject and within-subject variables."

The reason this baffles me is because I ran a similar syntax before with no problem, albeit I transformed this data into a team-level analysis, and so I no longer have an individual participant measure. I am trying to predict the likelihood for teams to select ideas that fit into three groups (categories that represent how much the idea diverges from the current paradigm of solving that problem). Below, please see my syntax from my PREVIOUS analysis at the individual level (just for main effects):

GENLIN Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeanNumber ParticipantNumber
(ORDER=ASCENDING) WITH IndividualKAI IndividualPsychSafety
/MODEL DesignTask Semester Instructor IndividualKAI IndividualPsychSafety
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=ParticipantNumber*TeanNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

Now for the NEW syntax manipulated to perform a similar analysis at the TEAM level. Note that we no longer have IndividualKAI, as that has been transformed into KAIDiversity and KAIElevation (standard dev. and mean measures for the team). We also have psychological safety as a team-level measure as PS_TP3:

GENLIN Final_Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeamNumber
(ORDER=ASCENDING) WITH KAIElevation KAIDiversity PS_TP3
/MODEL DesignTask Semester Instructor KAIElevation KAIDiversity PS_TP3
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

---

I've been trying to figure this out for a while. I thought maybe something was off with "REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count," but I tried to match the dataset file to what I had for my individual level analysis (the old one that worked). Any suggestions would be greatly appreciated!

------------------------------
Courtney Cole
------------------------------
#SPSSStatistics

Hello Rick!
I definitely agree that GEE would be ideal, but unfortunately GEE is more so "new territory" for the committee members I work with closely. So aggregating to a single percentage value was one of our ideas (hence potentially using hierarchical linear regression), where we would just have to change the research question slightly. I will play around with the data some more to see if I can solve the issue.

Thank you!

Courtney

------------------------------
Courtney Cole
------------------------------

Original Message:
Sent: Fri February 11, 2022 02:14 PM
From: Rick Marcantonio
Subject: Syntax issue? Same values for subject and within-subject variables?

From what little I know, if the behavior you are studying naturally has a within-ss component to it, then GEE would be the way to go. However, I'm no GEE expert, Courtney, so I can't say with any authority. Sounds like a question either for your advisor or someone on your committee who is familiar with your research and the precise hypothes(i|e)s you want to answer.

------------------------------
Rick Marcantonio
Quality Assurance
IBM

Original Message:
Sent: Fri February 11, 2022 02:06 PM
From: Courtney Cole
Subject: Syntax issue? Same values for subject and within-subject variables?

Hello Rick!
Thank you very much for your response. I tried running the syntax you posted for myself and removed Semester, so there is definitely some issue between TeamNumber and count. Is there a particular way I should have my within-subjects factor set up, or would you say GEE is not a viable option for my analysis? I know I can look at the selection process from the perspective of percentages of ideas selected by teams per each of the three categories (this would result in three analyses as a hierarchical linear regression), but if I wanted to look at my data more from the perspective of GEE or a multinomial logistic regression, I feel like my dataset as is would probably not work. I am okay with switching to hierarchical linear regression, but I just want to make sure I exhaust all my other options for the right reasons.

Thank you!

Courtney

------------------------------
Courtney Cole

Original Message:
Sent: Fri February 11, 2022 11:39 AM
From: Rick Marcantonio
Subject: Syntax issue? Same values for subject and within-subject variables?

Hi, Courtney. I tried this:

sort cases by TeamNumber Semester count.do if lag(TeamNumber)=TeamNumber and lag(Semester)=Semester and lag(count)=count.print / $casenum.end if.execute.​

and SPSS returned these case numbers as duplicates:

29, 33 35, 37, 40, 44. 47, 51, 53, 56, 61, 74, 80, 83, 90,
94, 99, 103, 112, 115, 118, 120, 122, 124, 127, 131, 139,
140, 144,  146, and 151.

So perhaps that's the problem....?

------------------------------
Rick Marcantonio
Quality Assurance
IBM

Original Message:
Sent: Thu February 10, 2022 01:57 PM
From: Courtney Cole
Subject: Syntax issue? Same values for subject and within-subject variables?

**FILES FOR BOTH DATASETS ATTACHED (RQ1 is the old file I am comparing the new file to)**

Hello all, I received the following error upon running my syntax: "There are at least two records with the same values for the subject and within-subject variables."

The reason this baffles me is because I ran a similar syntax before with no problem, albeit I transformed this data into a team-level analysis, and so I no longer have an individual participant measure. I am trying to predict the likelihood for teams to select ideas that fit into three groups (categories that represent how much the idea diverges from the current paradigm of solving that problem). Below, please see my syntax from my PREVIOUS analysis at the individual level (just for main effects):

GENLIN Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeanNumber ParticipantNumber
(ORDER=ASCENDING) WITH IndividualKAI IndividualPsychSafety
/MODEL DesignTask Semester Instructor IndividualKAI IndividualPsychSafety
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=ParticipantNumber*TeanNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

Now for the NEW syntax manipulated to perform a similar analysis at the TEAM level. Note that we no longer have IndividualKAI, as that has been transformed into KAIDiversity and KAIElevation (standard dev. and mean measures for the team). We also have psychological safety as a team-level measure as PS_TP3:

GENLIN Final_Rating_P (ORDER=ASCENDING) BY DesignTask Semester Instructor TeamNumber
(ORDER=ASCENDING) WITH KAIElevation KAIDiversity PS_TP3
/MODEL DesignTask Semester Instructor KAIElevation KAIDiversity PS_TP3
DISTRIBUTION=MULTINOMIAL
LINK=CUMLOGIT /CRITERIA METHOD=FISHER(1) SCALE=1
MAXITERATIONS=100 MAXSTEPHALVING=50 PCONVERGE=1E-006(ABSOLUTE)
SINGULAR=1E-012 ANALYSISTYPE=3(WALD) CILEVEL=95 LIKELIHOOD=KERNEL
/REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count SORT=YES CORRTYPE=EXCHANGEABLE
ADJUSTCORR=YES COVB=ROBUST MAXITERATIONS=100 PCONVERGE=1e-006(ABSOLUTE) UPDATECORR=1
/MISSING CLASSMISSING=EXCLUDE
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION WORKINGCORR.

---

I've been trying to figure this out for a while. I thought maybe something was off with "REPEATED SUBJECT=TeamNumber*Semester WITHINSUBJECT=count," but I tried to match the dataset file to what I had for my individual level analysis (the old one that worked). Any suggestions would be greatly appreciated!

------------------------------
Courtney Cole
------------------------------
#SPSSStatistics