RE: st: FW: ICC and loneway

["Mak, Timothy" <timothy.mak07@imperial.ac.uk>]

The ICC calculated by the -loneway- command assumes the two observers are exchangeable, ie that for each observation, you're recruiting two random person to evaluate the pain. If you only have two observers, you should want to take into account that there are just two people. Your ICC should be calculated differently. See McGraw and Wong (1996). Forming inferences about some intraclass correlation coefficients. Psychological Methods 1(1):30-46. I don't know of a Stata command that gives you the correct ICC, but perhaps others would know. Choosing the right ICC, however, is quite simple with SPSS. (One of the very few advantages of SPSS over Stata!)

Tim

 

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of JOHN ANTONAKIS
Sent: 15 January 2009 15:20
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: FW: ICC and loneway

You should have the data in the long format for this to work, with dummy 
(indicator) variable for x1 (=0) and x2(=1), not in the wide format as 
you do now. Type -help loneway- to see the example they use. To reshape 
the data look at the -reshape- command.

HTH,
John.

____________________________________________________

Prof. John Antonakis
Associate Dean
Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland

Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305

http://www.hec.unil.ch/people/jantonakis&cl=en
____________________________________________________



On 15.01.2009 13:56, Sham Lal wrote:
> Dear Statlisters
>
> I am looking to compare the ratings of pain between two observers x1 x2 and am using the loneway command to produce ICC's. However, I don't completely believe that loneway is producing what I require. My data is organised as follows:
>
> Idno    pain_x1     pain_x2
> A1        53                   74
> A2        62                   23
> B3        15                   12
> B5         86                  52  
>
> The output after running loneway pain_x1 pain_x2 is as follows:
> One-way Analysis of Variance for pain_x1: 
>
>                                               Number of obs =       203
>                                                   R-squared =    0.9025
>
>     Source                SS         df      MS            F     Prob > F
> -------------------------------------------------------------------------
> Between pain_x2         135346.15     68    1990.3846     18.25     0.0000
> Within pain_x2          14613.739    134    109.05775
> -------------------------------------------------------------------------
> Total                  149959.89    202     742.3757
>
>          Intraclass       Asy.        
>          correlation      S.E.       [95% Conf. Interval]
>          ------------------------------------------------
>             0.85839     0.03899       0.78197     0.93481
>
> This is comparing the variation between just pain_x2, I fear this is not correct?
> Any help would be appreciated
>
> Sham 
>
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>   
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/