Re: methodological question about difference scores

Bob Green (
Fri, 9 Apr 1999 06:04:18 +1000

Hello Travis,

Thanks for your reply. I have been away for Easter, hence my delay in
responding. Regarding your comments

>It rather depends on your precise question. For example, if you want to
>know whether or not the two people have the same structure amongst their
>personal constructs, you could obtain a KxK similarity matrix (using whatever
>similarity coefficient is warranted) for K constructs.

This is a sound point and raises the issue of how to examine issues such as
commonality and sociality. Upon reflection I am more interested at present
in whether two people could be considered to construe elements in a grid
similarly (in contrast to overall structure). Certainly MDS is a useful
option for examining structure, with weirdness and the subject weights being
useful. I have read about the application of canonical correlations though
am not really familiar with the approach and it is probably more complex
than what I am looking for.

The same may be said for QAP which I can't say I know anything about. It is
something I will consider. The 1983 reference did remind me of the
literature on interrater agreement (in the past I had used the ICC to
examine interrater agreement on some supplied constructs).

>Hubert, L. & Golledge, R.G., (1983). Rater agreement for complex
>assessments. British Journal of Mathematical and Statistical Psychology,
>36(2), 207-216

>Hope this helps,

Your comments are helpful, though perhaps not immediately so. What your
comments have done is assist me to focus on what was of interest to me,
namely, a relatively simple method-close to the raw data, which would
facilitate examining whether two people construe 2 or more elements
(dis)similarly. One means to do this via structural approaches as you
suggest, though I would like to explore the difference score and/or personal
projects analysis options further (mutual friction & mutual accord).



>>I would appreciate advice on calculating the difference between two grids.
>>I am not familar with the relevant literature though would appreciate
>>references or comments:
>>I am aware of two methods (though there may be other methods) to calculate
>>the level of agreement between two persons who rated the same elements on
>>the same constructs. These methods which do not give the same results are:
>>(1) A difference grid (the absolute difference for each cell, summed by row
>>and column)
>>(2) Transforming ratings into a "disagreement score", e.g., a 0 if there was
>>agreement, 1 if person
>>used 4 and the other 1 or 2, and 2 if they were on opposite ends of the
>>I also considered:
>>Bannister's consistency score
>>the possibility of joining the two grids and performing a cluster analysis
>>comparison on mean ratings for certain elements/constructs
>>Is there a methodologically preferable way of calculating difference (or
>>alternatively sameness).