[Sigia-l] obfuscating structure in card sorts?

[Eric Scheid]

I'm preparing a card sort exercise, and because I want to delve deeply into
the user's preferences for structure I'm supplying multiple instances and
examples of content nodes. Lots of cards.

Thus, rather than just one card that says "case study of method A", I have
six cards:
        
    case study of method A
    case study of method B
    case study of method C
    case study of method D
    case study of method E
    case study of method F

I also have six more cards, like this:

    cost/benefits analysis of method A
    cost/benefits analysis of method B
    cost/benefits analysis of method C
    cost/benefits analysis of method D
    cost/benefits analysis of method E
    cost/benefits analysis of method F

and another bunch of cards for each of the following...

    audience best served by method [A..F]
    example of client who used method [A..F]
    explanation of how method [A..F] is done
    pricing table for method [A..F]

and so on ... (*lots* of cards, I know, but that is a discussion for
another day)

It was at this point that it was really apparent that the wordings above
reflected quite strongly *my* mental structure for the information. It
wouldn't be surprising if some subjects arranged the cards into the above
structure at all, based only on the surface details of the artifacts.

It doesn't take much imagination to rewrite these cards using synonyms and
different phrase structures...

    case study of method A
    method B - a case study
    in-depth examination of a time we did method C
    method D - how we did it for a client, and the lessons learned
    extended anecdote involving method E
    etc

    cost/benefits analysis of method A
    method B - how the benefits outweigh the costs
    method C - what is the ROI?
    etc

Thus my personal preference and bias about how the information could/should
be organised is obfuscated and hopefully hidden from the test users.

Questions: Has anyone else been concerned that their own bias might unduly
influence the results? Has anyone done any card sorts similarly obfuscated
and noted a difference in results? Do you think the non-pattern wordings
encourage some subjects to actually think about what the card represents,
rather than simply arranging according to artifact details?

e.