Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: RE: Kappci problem


From   "Steichen, Thomas J." <[email protected]>
To   <[email protected]>
Subject   st: RE: RE: Kappci problem
Date   Thu, 21 Jun 2007 12:59:31 -0400

Let me restate my explanation for the second exampple...

I said "...the missing values are internally mis-coded to a "no" answer, thus leading to invalid stats."

However, for a string variable, "." is _not_ a missing value. Thus I should have said "...the periods (".") are internally coded to
a "no" answer, which likely was not intended, thus leading to invalid stats."

When a missing code (i.e., nothing) is substituted for each period, then this example causes the same error as the first example.

Tom

-----------------------------------
Thomas J. Steichen
[email protected]
-----------------------------------
 
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Steichen, Thomas J.
Sent: Thursday, June 21, 2007 12:00 PM
To: [email protected]
Subject: st: RE: Kappci problem


In the first example, -var3- is constant, thus an expected internal 2x2 matrix becomes a 2x1 matrix and locals assigned to the value
of the missing column become missing instead of zero, causing the error message.

In the second example, the missing values are internally mis-coded to a "no" answer, thus leading to invalid stats.

Simply put, as coded, kappci will only function correctly when there are no missing values and no variable is constant.

Tom

-----------------------------------
Thomas J. Steichen
[email protected]
-----------------------------------
 
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Nigel Armfield
Sent: Thursday, June 21, 2007 7:29 AM
To: [email protected]
Cc: Mark Bensink
Subject: st: Kappci problem

Hi,

I am trying to an analysis of agreement between two raters (dichotomous
variables) using kappci.
Each set of data is 14 pairs of Y/N.

I'm confused by the behaviour of kappci - sometimes it produces a result
that looks fine, sometimes it produces a result but the kappa value
seems way too low. Other times it fails with an odd error message.

Here's an example of the error:

var1	var2	var3
1	y	y
2	y	y
3	n	y
4	y	y
5	y	y
6	y	y
7	y	y
8	y	y
9	y	y
10	n	y
11	y	y
12	y	y
13	y	y
14	y	y

.. kappci var2 var3, pos("y")
'.' found where number expected
r(7);

I have re-entered the data manually but with no luck.

With a different set of data:

var1	var2	var3
1	y	y
2	n	y
3	y	y
4	y	y
5	y	y
6	y	y
7	y	y
8	y	y
9	y	y
10	y	y
11	y	y
12	.	.
13	y	y
14	y	y

.. kappci var2 var3,pos("y")

 Expected    Observed   -- Binomial Exact --
Agreement   Agreement   [95% Conf. Interval]
--------------------------------------------
   80.61%      92.86%      66.13%     99.82%

  Kappa  [95% Conf. Interval]        Z     Prob>Z
-------------------------------------------------
 0.6316     0.0418     0.9325     2.54     0.0055

Any ideas would be gratefully received - I've been going round in
circles and getting nowhere fast for the last few hours.

Thanks in advance,

Nigel

--

Nigel R. Armfield. MSc, MIEEE, MIET.
Senior Research Assistant (Telepaediatrics)
Centre for Online Health
The University of Queensland
Level 3, Royal Children's Hospital
Herston 4029
Australia

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

-----------------------------------------
CONFIDENTIALITY NOTE: This e-mail message, including any
attachment(s), contains information that may be confidential,
protected by the attorney-client or other legal privileges, and/or
proprietary non-public information. If you are not an intended
recipient of this message or an authorized assistant to an intended
recipient, please notify the sender by replying to this message and
then delete it from your system. Use, dissemination, distribution,
or reproduction of this message and/or any of its attachments (if
any) by unintended recipients is not authorized and may be
unlawful.


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index