Re: repeated measure anova error

[Art Kendall]

The GUI (graphic user interface) is a great way to get the first draft of your syntax.

Go through your clicks in the interface and exit via <paste>.
You can then copy that syntax into a mail message.

Also please describe what you are trying to do.   How many and what between subjects factors do you have if any?  How many and what within subjects factors?  How many levels in each factor?  Do you only have two repeats of the DV for each subject?


What do you get when you do MEANS on your DVs by your between subjects factors?


Art Kendall
Social Research Consultants

On 2/20/2010 3:20 PM, Eva Glieberman wrote:

I'm honestly fairly new to SPSS, and I'm not using syntax, but rather using the command interface features.  I select the run repeated measures option from the analysis/general linear model tabs.  Then I plugged in my dependent variables for the factors and binary independent variables for the between subject option.

On Sat, Feb 20, 2010 at 3:17 PM, Art Kendall <[hidden email]> wrote:

What syntax are you using to do this?

Art Kendall
Social Research Consultants

On 2/20/2010 2:45 PM, Eva wrote:

I keep running across this error when I try to run my SPSS model: "no
valid cases found.  this command cannot be executed."

I am told this is probably because I didn't merge or code my variables
properly.  How do I track where my error is?

Thanks.  I can post more output if helpful.
 

===================== 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

[ViAnn Beadle]

I would check for cases deleted because of missing values. If listwise deletion is in effect (usually the default), a case is deleted from the analysis if any variable specified in the model is missing for that case.

On Sat, Feb 20, 2010 at 2:38 PM, Art Kendall <[hidden email]> wrote:

The GUI (graphic user interface) is a great way to get the first draft of your syntax.

Go through your clicks in the interface and exit via <paste>.
You can then copy that syntax into a mail message.

Also please describe what you are trying to do.   How many and what between subjects factors do you have if any?  How many and what within subjects factors?  How many levels in each factor?  Do you only have two repeats of the DV for each subject?


What do you get when you do MEANS on your DVs by your between subjects factors?


Art Kendall
Social Research Consultants

On 2/20/2010 3:20 PM, Eva Glieberman wrote:

I'm honestly fairly new to SPSS, and I'm not using syntax, but rather using the command interface features.  I select the run repeated measures option from the analysis/general linear model tabs.  Then I plugged in my dependent variables for the factors and binary independent variables for the between subject option.

On Sat, Feb 20, 2010 at 3:17 PM, Art Kendall <[hidden email]> wrote:

What syntax are you using to do this?

Art Kendall
Social Research Consultants

On 2/20/2010 2:45 PM, Eva wrote:

I keep running across this error when I try to run my SPSS model: "no
valid cases found.  this command cannot be executed."

I am told this is probably because I didn't merge or code my variables
properly.  How do I track where my error is?

Thanks.  I can post more output if helpful.
 

===================== 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

[Art Kendall]

It is better if you include the newsgroup and/or list in your posts. 1) It builds up the archives so that people with a similar problem in the future can find it. 2) more people will see your posts and the replies and can add more suggestions 3) the person who replies may not get back to the newsgroup/list as quickly as other members.

Check which dataset should be activated. click <windows> and look at the contents of the different datasets you have open.

Art Kendall
Social Research Consultants


On 2/20/2010 8:17 PM, Eva Glieberman wrote:

Thanks for the reply.  Sorry for the lack of clarity.  I am currently trying to write my master's thesis in public health.  The topic is on the combined impact of child abuse and attention deficit disorder on childhood aggressive outcomes.  I am trying to run a repeated measures ANOVA.  My ind. variables are child abuse at age 4 and ADHD at age 4 (as defined by measures used in the field) and my dependent variable is aggression (with four levels).

 

When I try to run means, I get an error message--"no valid cases were found."

 

The code that I've plugged in as follows:

 

DATASET ACTIVATE DataSet2. DATASET CLOSE DataSet1. GLM cb4tagg4c cb4tagg6c cb4tagg8c cb4tagg10c BY cb4tatt4 cb4tagg4 /WSFACTOR=factor1 4 Polynomial /METHOD=SSTYPE(3) /EMMEANS=TABLES(cb4tatt4*cb4tagg4*factor1) /PRINT=DESCRIPTIVE /CRITERIA=ALPHA(.05) /WSDESIGN=factor1 /DESIGN=cb4tatt4 cb4tagg4 cb4tatt4*cb4tagg4.

 

-Eva

On Sat, Feb 20, 2010 at 4:38 PM, Art Kendall <[hidden email]> wrote:

The GUI (graphic user interface) is a great way to get the first draft of your syntax.

Go through your clicks in the interface and exit via <paste>.
You can then copy that syntax into a mail message.

Also please describe what you are trying to do.   How many and what between subjects factors do you have if any?  How many and what within subjects factors?  How many levels in each factor?  Do you only have two repeats of the DV for each subject?


What do you get when you do MEANS on your DVs by your between subjects factors?

Art Kendall
Social Research Consultants

On 2/20/2010 3:20 PM, Eva Glieberman wrote:

I'm honestly fairly new to SPSS, and I'm not using syntax, but rather using the command interface features.  I select the run repeated measures option from the analysis/general linear model tabs.  Then I plugged in my dependent variables for the factors and binary independent variables for the between subject option.

On Sat, Feb 20, 2010 at 3:17 PM, Art Kendall <[hidden email]> wrote:

What syntax are you using to do this?

Art Kendall
Social Research Consultants

On 2/20/2010 2:45 PM, Eva wrote:

I keep running across this error when I try to run my SPSS model: "no
valid cases found.  this command cannot be executed."

I am told this is probably because I didn't merge or code my variables
properly.  How do I track where my error is?

Thanks.  I can post more output if helpful.
 

===================== 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

[Mike]

If I can butt in: Just to clarify, you claim that you have two (quasi-)independent

variables which are:

cb4tatt4   (a measure of attention deficit at age 4?)

cb4tagg4  (a measure of abuse at age 4?

 

and four "dependent variables" which are four different measures

of aggression:

cb4tagg4c

cb4tagg6c

cb4tagg8c

cb4tagg10c

 

You seem to be imposing a THIRD independent variable on the four

measures of aggression, what you refer to as "factor1".  You need

to be clearer on whether you have a 2x2x4 Mixed design (i.e., 2 between-subjects

factors and 1 within-subject factor) or a 2x2 completely between-subjects

designs with 4 dependent variables (which would suggest you want to

do a MANOVA analysis).

 

In any event, what do you get if you run the following code:

 

Means tables=cb4tagg4c cb4tagg6c cb4tagg8c cb4tagg10c

           BY cb4tatt4 by cb4tagg4.

 

The above procedure should produce a single table for each of your

four dependent variables divided into subgroups defined by

the 2x2 design consisting of the variables "cb4tatt4 by cb4tagg4".

 

I think the above might provides what you specifying in your

Emmeans statement below.

 

You can also just check your data to make sure that you have

any missing values for the six variables you're using, that is:

 

list var=ID cb4tatt4 by cb4tagg4 cb4tagg4c cb4tagg6c cb4tagg8c cb4tagg10c.

 

It could be that case that you have a lot of missing values.  If it

turns out that you have few missing values, then I think the problem

may be incorrect design specification to the SPSS procedures.

 

-Mike Palij

New York University

[hidden email]

 

          

 

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Tuesday, February 23, 2010 8:02 AM

Subject: Re: repeated measure anova error

It is better if you include the newsgroup and/or list in your posts. 1) It builds up the archives so that people with a similar problem in the future can find it. 2) more people will see your posts and the replies and can add more suggestions 3) the person who replies may not get back to the newsgroup/list as quickly as other members.

Check which dataset should be activated. click <windows> and look at the contents of the different datasets you have open.

Art Kendall
Social Research Consultants


On 2/20/2010 8:17 PM, Eva Glieberman wrote:

Thanks for the reply.  Sorry for the lack of clarity.  I am currently trying to write my master's thesis in public health.  The topic is on the combined impact of child abuse and attention deficit disorder on childhood aggressive outcomes.  I am trying to run a repeated measures ANOVA.  My ind. variables are child abuse at age 4 and ADHD at age 4 (as defined by measures used in the field) and my dependent variable is aggression (with four levels).

 

When I try to run means, I get an error message--"no valid cases were found."

 

The code that I've plugged in as follows:

 

DATASET ACTIVATE DataSet2. DATASET CLOSE DataSet1. GLM cb4tagg4c cb4tagg6c cb4tagg8c cb4tagg10c BY cb4tatt4 cb4tagg4 /WSFACTOR=factor1 4 Polynomial /METHOD=SSTYPE(3) /EMMEANS=TABLES(cb4tatt4*cb4tagg4*factor1) /PRINT=DESCRIPTIVE /CRITERIA=ALPHA(.05) /WSDESIGN=factor1 /DESIGN=cb4tatt4 cb4tagg4 cb4tatt4*cb4tagg4.

 

-Eva

On Sat, Feb 20, 2010 at 4:38 PM, Art Kendall <[hidden email]> wrote:

The GUI (graphic user interface) is a great way to get the first draft of your syntax.

Go through your clicks in the interface and exit via <paste>.
You can then copy that syntax into a mail message.

Also please describe what you are trying to do.   How many and what between subjects factors do you have if any?  How many and what within subjects factors?  How many levels in each factor?  Do you only have two repeats of the DV for each subject?


What do you get when you do MEANS on your DVs by your between subjects factors?

Art Kendall
Social Research Consultants

On 2/20/2010 3:20 PM, Eva Glieberman wrote:

I'm honestly fairly new to SPSS, and I'm not using syntax, but rather using the command interface features.  I select the run repeated measures option from the analysis/general linear model tabs.  Then I plugged in my dependent variables for the factors and binary independent variables for the between subject option.

On Sat, Feb 20, 2010 at 3:17 PM, Art Kendall <[hidden email]> wrote:

What syntax are you using to do this?

Art Kendall
Social Research Consultants

On 2/20/2010 2:45 PM, Eva wrote:

I keep running across this error when I try to run my SPSS model: "no
valid cases found.  this command cannot be executed."

I am told this is probably because I didn't merge or code my variables
properly.  How do I track where my error is?

Thanks.  I can post more output if helpful.
 

===================== 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