There are many places in SPSS where one can get some of the many kinds of ICCs. The algorithms are available in SPSS under <help>
I suggest that you -- Google "nabble SPSSX" --join that discussion group which has many very experienced statisticians --post your query with the syntax you are using and a description of you data. What are the variables? What is their level of measurement? What is the response scale? How many scorers? How many times did a scorer see the same scan? How many scans are there?
How different are the results?
If the data is not extensive and is not confidential you might also attach your data.
As a courtesy to this list, post what you find out back on this list.
Art Kendall Social Research Consultants
On 5/18/2013 4:22 AM, email@example.com wrote: > Hi there, > > I am calculating intraclass correlation coefficients to assess the reliability of visual scoring of brain scans in Parkinson's. > > However, I did this three times: > *in SPSS, > *in software available from the university of Hong Kong at http://department.obg.cuhk.edu.hk/researchsupport/IntraClass_correlation.asp > *and then in Excel using the formulas given on http://luna.cas.usf.edu/~mbrannic/files/pmet/shrout1.htm > > The results I got from the calculation 'by hand' were exactly equal to those delivered by the Hong Kong website. To my amazement, however, these differ from the SPSS results. > Could anyone explain what is happening here? > > Thanks, > > Frank >