ANOVA and Regression question with summary Likert scale data

[00000506dfb4cdf3-dmarc-request]

Are there any special considerations in ANOVA and regression SPSS procedures with using summary data, i.e. means of survey items (ordinal data) as opposed to raw data? Of course there is the question whether such procedures should be used at all with ordinal data, but other than that, any thoughts, There are no provisions in the menus for such options. If there is a case for summary data considerations, does anyone have syntax to achieve this? Thank you in advance. Ralph

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

[David Marso]

In the FM you will find the MATRIX DATA command and within various procedures a MATRIX IN subcommand.  I leave it to you to research this avenue as I have no syntax readily at hand.
To get an idea of exactly what is expected for various procedures you can use MATRIX OUT (<filespec>) and then study the resulting dataset. <filespec> can be either * (active file), a qualified datapath, or a declared dataset.

00000506dfb4cdf3-dmarc-request wrote

Are there any special considerations in ANOVA and regression SPSS  
procedures with using summary data, i.e. means of survey items (ordinal  data) as
opposed to raw data? Of course there is the question whether such  procedures
should be used at all with ordinal data, but other than that, any  thoughts,
There are no provisions in the menus for such options. If there is a  case
for summary data considerations, does anyone have syntax to achieve this?  
Thank you in advance. Ralph

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

[Maguin, Eugene]

Adding to David's post. You have to use the Manova command as that command is the only Anova-type command that can process summary data and that summary data has to be presented as a matrix. As you may know, GLM can be used to do regressions and I suspect, but have never tried to do so, that Manova also can be used to do regressions.  

Here is a repeated measures example.

matrix data
 variables = group rowtype_ dv1 dv2 dv3
 /factor = group
 /format = lower nodiagonal.
begin data
1 mean 0.00 0.21 0.42
1 n 100 100 100
2 mean 0.00 0.21 0.42
2 n 100 100 100
3 mean 0.00 0.00 0.00
3 n 100 100 100
. sd 1.0 1.0 1.0
. corr 0.4
. corr 0.4 0.4
end data.

manova dv1 dv2 dv3 by group(1,3)/wsfactors dv(3)/method = unique
 /error = within+residual/matrix = in(*)/power t (.05) F (.05)
 /print signif (mult averf)/noprint param(estim).


Gene Maguin


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of David Marso
Sent: Thursday, February 04, 2016 3:42 PM
To: [hidden email]
Subject: Re: ANOVA and Regression question with summary Likert scale data

In the FM you will find the MATRIX DATA command and within various
procedures a MATRIX IN subcommand.  I leave it to you to research this
avenue as I have no syntax readily at hand.
To get an idea of exactly what is expected for various procedures you can
use MATRIX OUT (<filespec>) and then study the resulting dataset. <filespec>
can be either * (active file), a qualified datapath, or a declared dataset.


00000506dfb4cdf3-dmarc-request wrote

> Are there any special considerations in ANOVA and regression SPSS  
> procedures with using summary data, i.e. means of survey items (ordinal
> data) as
> opposed to raw data? Of course there is the question whether such
> procedures
> should be used at all with ordinal data, but other than that, any
> thoughts,
> There are no provisions in the menus for such options. If there is a  case
> for summary data considerations, does anyone have syntax to achieve this?  
> Thank you in advance. Ralph
>
> =====================
> 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





-----
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?"
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/ANOVA-and-Regression-question-with-summary-Likert-scale-data-tp5731439p5731440.html
Sent from the SPSSX Discussion mailing list archive at 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

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

[David Marso]

ONEWAY and REGRESSION also support MATRIX DATA input.

DATA LIST FREE /GP Y .
BEGIN DATA
1 20  1 34  1 23  1  4  1 7
2 10  2  5  2 10  2 12  2 10
3  1  3  7  3  9  3 12  3 10
END DATA.
DATASET NAME raw.


DATASET DECLARE reg.
REGRESSION / MATRIX OUT (reg)/VARIABLES GP Y / DEPENDENT Y / METHOD ENTER GP .
DATASET ACTIVATE reg.
REGRESSION MATRIX IN (reg) /VARIABLES GP Y / DEPENDENT Y / METHOD ENTER GP .
LIST.

DATASET ACTIVATE raw.
DATASET DECLARE oneway.
ONEWAY Y BY GP / MATRIX OUT (oneway).
ONEWAY Y BY GP /MATRIX IN (oneway) .

Maguin, Eugene wrote

Adding to David's post. You have to use the Manova command as that command is the only Anova-type command that can process summary data and that summary data has to be presented as a matrix. As you may know, GLM can be used to do regressions and I suspect, but have never tried to do so, that Manova also can be used to do regressions.  

Here is a repeated measures example.

matrix data
 variables = group rowtype_ dv1 dv2 dv3
 /factor = group
 /format = lower nodiagonal.
begin data
1 mean 0.00 0.21 0.42
1 n 100 100 100
2 mean 0.00 0.21 0.42
2 n 100 100 100
3 mean 0.00 0.00 0.00
3 n 100 100 100
. sd 1.0 1.0 1.0
. corr 0.4
. corr 0.4 0.4
end data.

manova dv1 dv2 dv3 by group(1,3)/wsfactors dv(3)/method = unique
 /error = within+residual/matrix = in(*)/power t (.05) F (.05)
 /print signif (mult averf)/noprint param(estim).


Gene Maguin


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of David Marso
Sent: Thursday, February 04, 2016 3:42 PM
To: [hidden email]
Subject: Re: ANOVA and Regression question with summary Likert scale data

In the FM you will find the MATRIX DATA command and within various
procedures a MATRIX IN subcommand.  I leave it to you to research this
avenue as I have no syntax readily at hand.
To get an idea of exactly what is expected for various procedures you can
use MATRIX OUT (<filespec>) and then study the resulting dataset. <filespec>
can be either * (active file), a qualified datapath, or a declared dataset.


00000506dfb4cdf3-dmarc-request wrote
> Are there any special considerations in ANOVA and regression SPSS  
> procedures with using summary data, i.e. means of survey items (ordinal
> data) as
> opposed to raw data? Of course there is the question whether such
> procedures
> should be used at all with ordinal data, but other than that, any
> thoughts,
> There are no provisions in the menus for such options. If there is a  case
> for summary data considerations, does anyone have syntax to achieve this?  
> Thank you in advance. Ralph
>
> =====================
> 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





-----
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?"
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/ANOVA-and-Regression-question-with-summary-Likert-scale-data-tp5731439p5731440.html
Sent from the SPSSX Discussion mailing list archive at 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

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

[Maguin, Eugene]

I missed checking Oneway and looked in the wrong place on Regression. Thanks for correcting my error. Gene Maguin

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of David Marso
Sent: Thursday, February 04, 2016 4:32 PM
To: [hidden email]
Subject: Re: ANOVA and Regression question with summary Likert scale data

ONEWAY and REGRESSION also support MATRIX DATA input.

DATA LIST FREE /GP Y .
BEGIN DATA
1 20  1 34  1 23  1  4  1 7
2 10  2  5  2 10  2 12  2 10
3  1  3  7  3  9  3 12  3 10
END DATA.
DATASET NAME raw.


DATASET DECLARE reg.
REGRESSION / MATRIX OUT (reg)/VARIABLES GP Y / DEPENDENT Y / METHOD ENTER GP
.
DATASET ACTIVATE reg.
REGRESSION MATRIX IN (reg) /VARIABLES GP Y / DEPENDENT Y / METHOD ENTER GP .
LIST.

DATASET ACTIVATE raw.
DATASET DECLARE oneway.
ONEWAY Y BY GP / MATRIX OUT (oneway).
ONEWAY Y BY GP /MATRIX IN (oneway) .


Maguin, Eugene wrote

> Adding to David's post. You have to use the Manova command as that command
> is the only Anova-type command that can process summary data and that
> summary data has to be presented as a matrix. As you may know, GLM can be
> used to do regressions and I suspect, but have never tried to do so, that
> Manova also can be used to do regressions.  
>
> Here is a repeated measures example.
>
> matrix data
>  variables = group rowtype_ dv1 dv2 dv3
>  /factor = group
>  /format = lower nodiagonal.
> begin data
> 1 mean 0.00 0.21 0.42
> 1 n 100 100 100
> 2 mean 0.00 0.21 0.42
> 2 n 100 100 100
> 3 mean 0.00 0.00 0.00
> 3 n 100 100 100
> . sd 1.0 1.0 1.0
> . corr 0.4
> . corr 0.4 0.4
> end data.
>
> manova dv1 dv2 dv3 by group(1,3)/wsfactors dv(3)/method = unique
>  /error = within+residual/matrix = in(*)/power t (.05) F (.05)
>  /print signif (mult averf)/noprint param(estim).
>
>
> Gene Maguin
>
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:

> SPSSX-L@.UGA

> ] On Behalf Of David Marso
> Sent: Thursday, February 04, 2016 3:42 PM
> To:

> SPSSX-L@.UGA

> Subject: Re: ANOVA and Regression question with summary Likert scale data
>
> In the FM you will find the MATRIX DATA command and within various
> procedures a MATRIX IN subcommand.  I leave it to you to research this
> avenue as I have no syntax readily at hand.
> To get an idea of exactly what is expected for various procedures you can
> use MATRIX OUT (
> <filespec>
> ) and then study the resulting dataset.
> <filespec>
> can be either * (active file), a qualified datapath, or a declared
> dataset.
>
>
> 00000506dfb4cdf3-dmarc-request wrote
>> Are there any special considerations in ANOVA and regression SPSS  
>> procedures with using summary data, i.e. means of survey items (ordinal
>> data) as
>> opposed to raw data? Of course there is the question whether such
>> procedures
>> should be used at all with ordinal data, but other than that, any
>> thoughts,
>> There are no provisions in the menus for such options. If there is a
>> case
>> for summary data considerations, does anyone have syntax to achieve this?  
>> Thank you in advance. Ralph
>>
>> =====================
>> 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
>
>
>
>
>
> -----
> 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?"
> --
> View this message in context:
> http://spssx-discussion.1045642.n5.nabble.com/ANOVA-and-Regression-question-with-summary-Likert-scale-data-tp5731439p5731440.html
> Sent from the SPSSX Discussion mailing list archive at 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
>
> =====================
> 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





-----
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?"
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/ANOVA-and-Regression-question-with-summary-Likert-scale-data-tp5731439p5731442.html
Sent from the SPSSX Discussion mailing list archive at 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

=====================
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]

Do you mean that your cases are summaries?  or do you mean that you have a set of  type items that are intended as items in a summative scale?

Sometimes people use "ordinal" item response scale loosely?  What is your item response scale?
What is the rational for considering the items severely discrepant from interval level?

If it is a summative scale is it one that has had reliability established by prior research?

if not how did you come up with the stimulus questions?

It it is a summative scale it is rarely necessary to treat the mean or sum as ordinal. (in most instances even a single item with equal appearing intervals such as a Likert item does not show any meaningful difference when ordinal vs interval assumption are tested in the CATREG procedure.)