Analyse voting data in small groups

[Lari Ono]

What is the best statistic method to analyse voting data about some defined
subject in small groups?

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

[Peter Spangler]

What is your hypothesis to test? This is a good place to start

Sent from my iPhone

On Mar 1, 2013, at 8:46 AM, Lari Ono <[hidden email]> wrote:

> What is the best statistic method to analyse voting data about some defined
> subject in small groups?
>
> =====================
> 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

[Lari Ono]

The hypothesis is " what attributes (approximately 15) are relevant to
determine if the vote is positive or negative".

The data are decisions on articles to be accepted in events.

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

[Peter Spangler]

By attributes do you mean the response levels of your different
variables? Not sure what your variables (columns) are, continuous,
nominal ordinal?

Sent from my iPhone

On Mar 1, 2013, at 9:28 AM, Lari Ono <[hidden email]> wrote:

> The hypothesis is " what attributes (approximately 15) are relevant to
> determine if the vote is positive or negative".
>
> The data are decisions on articles to be accepted in events.
>
> =====================
> 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]

Perhaps you would care to elaborate a little bit (or a lot)?
This is not twitter.  
Please feel free to describe your data and collection methods in sufficient detail to make it possible for someone to provide a reasonable answer!
--

Lari Ono wrote

The hypothesis is " what attributes (approximately 15) are relevant to
determine if the vote is positive or negative".

The data are decisions on articles to be accepted in events.

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

[Rich Ulrich]

Delphi polling.

--
Rich Ulrich

> Date: Fri, 1 Mar 2013 11:43:36 -0500
> From: [hidden email]
> Subject: Analyse voting data in small groups
> To: [hidden email]
>
> What is the best statistic method to analyse voting data about some defined
> subject in small groups?
>
...

[Lari Ono]

Sorry for the little bit information.

My data has the following attributes:

Name Type Range of values
Vote Nominal {0,1,0.5}
Final Decision Nominal {0,1,0.5}
Vote win Ordinal {0,1}
"Ideology" Scale {-1...1}
Amount of positive votes Ordinal {0,1}
Amount of negative votes Ordinal {0,1}
Amount of parcial votes Ordinal {0,1}
Amount of votes Ordinal {0,1}
State Author Nominal {0...25}
State Revisor Nominal {0...25}
Scholarity Nominal {0...4}
Gender Nominal {0,1}
College Revisor Nominal {0..15}
Birth year Ordinal {1950..1980}
Average gender peers Scale {0...1}
Average "ideology" peers Scale {0...1}
Average birth year peers Scale {0...1}
Average vote peers Scale {0...1}


I got this information from various events at my university, which evaluated
articles.

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

First go back and review the levels of measurement.
Dichotomies (variables that can take only 2 values) are special.  They can be taken as interval.
Counts are interval level.

I suggest that you go back and fix your data view.  Include more information in the label for each variable, e.g., what do you mean by "State author", "State revisor" "Partial vote" etc. Then post the dictionary including value labels.

Also please include a fuller description of your procedure, and what questions you are trying to answer.


On 3/2/2013 7:50 AM, Lari Ono wrote:

Sorry for the little bit information.

My data has the following attributes:

Name Type Range of values
Vote Nominal {0,1,0.5}
Final Decision Nominal {0,1,0.5}
Vote win Ordinal {0,1}
"Ideology" Scale {-1...1}
Amount of positive votes Ordinal {0,1}
Amount of negative votes Ordinal {0,1}
Amount of parcial votes Ordinal {0,1}
Amount of votes Ordinal {0,1}
State Author Nominal {0...25}
State Revisor Nominal {0...25}
Scholarity Nominal {0...4}
Gender Nominal {0,1}
College Revisor Nominal {0..15}
Birth year Ordinal {1950..1980}
Average gender peers Scale {0...1}
Average "ideology" peers Scale {0...1}
Average birth year peers Scale {0...1}
Average vote peers Scale {0...1}


I got this information from various events at my university, which evaluated
articles.

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