Using NSKAPPA MACRO

[Placide-2]

Dear All,

I am using NSKAPPA MACRO provided by SPSS

http://support.spss.com/ProductsExt/Statistics/Documentation/Statistics/macros/Nskappa.htm

I got a division by zero error message when computing Kappa for two raters (see explanations hereunder).

Thank you for your help

Placide

 

My case is:

Values: 1= TEC; 2=HUM; 3=SYU;  .........; 9=DSNTFT

Rater1	Rater2
TEC	TEC
TEC	SYU
TEC	SYU
TEC	TEC
TEC	SYU

 

 

[Bruce Weaver]

Placide-2 wrote

Dear All,
I am using NSKAPPA MACRO provided by SPSS
http://support.spss.com/ProductsExt/Statistics/Documentation/Statistics/macros/Nskappa.htm
I got a division by zero error message when computing Kappa for two raters (see explanations hereunder).
Thank you for your help
Placide
 
My case is:
Values: 1= TEC; 2=HUM; 3=SYU;  .........; 9=DSNTFT

--- snip table ---

I want to make sure I understand your "case".  Do you mean you have a 3x3 table (rater 1 x rater 2) with 5 observations in total, 2 in the top left cell, and 3 in the top right cell?

[Placide-2]

Dear All,

I have 2 raters assigning one values from 1 to 9 (1= TEC; 2=HUM; 3=SYU; .......; 9=DSNTFT) to each of my 5 subjects.

The Macro is generating a "division by zero" error dispite the fact that it is suppose to compute Kappa for non squared CrossTab

Thank you for your Help

Placide

--- On Wed, 2/10/10, Bruce Weaver <[hidden email]> wrote:

From: Bruce Weaver <[hidden email]>
Subject: Re: Using NSKAPPA MACRO
To: [hidden email]
Date: Wednesday, February 10, 2010, 3:38 PM

Placide-2 wrote:

>
> Dear All,
> I am using NSKAPPA MACRO provided by SPSS
> http://support.spss.com/ProductsExt/Statistics/Documentation/Statistics/macros/Nskappa.htm
> I got a division by zero error message when computing Kappa for two raters
> (see explanations hereunder).
> Thank you for your help
> Placide
> Â
> My case is:
> Values: 1= TEC; 2=HUM; 3=SYU; Â .........; 9=DSNTFT
>
> --- snip table ---
>

I want to make sure I understand your "case".  Do you mean you have a 3x3
table (rater 1 x rater 2) with 5 observations in total, 2 in the top left
cell, and 3 in the top right cell?



-----
--
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.
--
View this message in context: http://old.nabble.com/Using-NSKAPPA-MACRO-tp27536323p27538537.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@... (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]

Your table has 81 cells.  5 subjects * 2 raters ==> 10 cells.   10 out of 81 is a very sparse matrix.  I doubt any summarization of that table would be meaningful.  WRT the table you posted,  in going back for the redesign you only know: rater1 thought everybody of the 5 subjects was "TEC" and (s)he used no other categories. Rater 2 used only 2 categories of the none. Before you go on you might want to debrief your raters about their understanding of the measurement scheme.

The first rater had no variability so it is impossible for her/his ratings to covary (correlate with anything).  There is no covariance to measure.

As you redesign your studies you will want to use subjects you have reason to believe vary in their "true" score as part of your pretesting.

Art Kendall
Social Research Consultants.

On 2/11/2010 10:13 AM, placide poba-nzaou wrote:

Dear All,

I have 2 raters assigning one values from 1 to 9 (1= TEC; 2=HUM; 3=SYU; .......; 9=DSNTFT) to each of my 5 subjects.

The Macro is generating a "division by zero" error dispite the fact that it is suppose to compute Kappa for non squared CrossTab

Thank you for your Help

Placide

--- On Wed, 2/10/10, Bruce Weaver [hidden email] wrote:

From: Bruce Weaver [hidden email]
Subject: Re: Using NSKAPPA MACRO
To: [hidden email]
Date: Wednesday, February 10, 2010, 3:38 PM

Placide-2 wrote:
>
> Dear All,
> I am using NSKAPPA MACRO provided by SPSS
> http://support.spss.com/ProductsExt/Statistics/Documentation/Statistics/macros/Nskappa.htm
> I got a division by zero error message when computing Kappa for two raters
> (see explanations hereunder).
> Thank you for your help
> Placide
> Â
> My case is:
> Values: 1= TEC; 2=HUM; 3=SYU; Â .........; 9=DSNTFT
>
> --- snip table ---
>

I want to make sure I understand your "case".  Do you mean you have a 3x3
table (rater 1 x rater 2) with 5 observations in total, 2 in the top left
cell, and 3 in the top right cell?



-----
--
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.
--
View this message in context: http://old.nabble.com/Using-NSKAPPA-MACRO-tp27536323p27538537.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@... (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