[Sigia-l] conflict of interest (was: usability testing)

[Arthur Fink]

Conflict of interest is a structural relationship that might 
reasonably be expected to provide either the reality or the 
appearance of a pressure -- however subtle -- to act in a certain way.

This is the "classical" case:

* I serve on a town board, which is considering a zoning issue for an 
applicant who happens to be a major client of mine.  That's 
considered a conflict of interest, and I was be expected to recuse 
myself, because I might be expected to decide in favor of my client, 
lest I upset him and lose his business.  It doesn't matter that I 
personally feel completely capable of deciding the case on its merits.

I actually had a prospect to whom I had submitted a proposal, ask for 
help in evaluating my proposal and the other competing ones.  I felt 
compelled to tell him that he could either use me as a consultant or 
consider me as one of the applicants.  I couldn't trust myself to 
honestly evaluate my proposal against the others (even though I 
certainly fancied that I could!).

But I doubt that this is what matters to most of us.  Here's an issue 
closer to home:  Could I design a UI and iplement that UI?  Sure --I 
could.  But would my design be biased so that my implementation would 
be easier?  (Which, I must say, might not be a bad thing.)  Or, would 
my attention to some of the complex implementation issues distract me 
from seeing bigger usablity concerns.  The "macho" in me wants to 
say, "no problems".  The realist notes that these are real concerns, 
and that I would be most comfortable, and offer the most service to 
my client, by just doing the design part, and doing it very well.

Having structural ties, biases, limited range of vision, and such are 
not necessarily bad things.  What's important is simply that we 
recognize them, frolic in our own frailty, and then get to work.

Fine doctors are used to having their clients seek second or third 
opinions.  When offered the prospect of a serious surgical procedure, 
or the lack of one when some would suggest it's needed, I'd be 
inclined to engage another doctor to look over the whole decision 
process, the test, etc -- even if I had complete faith in the 
integrity and skill of the first doctor.  And when the stakes to my 
health are so high --why not?  Well, can we be on the other side.

If there's trust and openness, recognition that testing can be part 
of a iterative design process or just a validation effort, and a 
genuine desire to serve the client with integrity -- what's the problem?

Arthur Fink

At 02:42 AM 1/28/2007, jamie foggon wrote:
>Ziya Oz:
>
> > For the last time: I've said "Conflict of interest *is* bias" and NOT "Bias
> > is conflict of interest."
> >
>
>Wikipedia's 'Bias' entry:
>
>"one is generally only said to be biased if one's powers of judgment
>are influenced by the biases one holds, to the extent that one's views
>could not be taken as being neutral or objective"
>
>My take on it:
>
>Having a bias is not something that can be measured, it just means
>that someone's viewpoint or behaviour cannot be taken as neutral. The
>only way bias could be separated from conflict of interest is if the
>degree of bias could be mesaured quantitatively. Then you could say
>person x was in a conflict of interest position, but this resulted in
>0% bias when they made decisions.
>
>
>--
>Jamie Foggon
>------------
>IA Summit 2007:  Enriching IA
>Rich Information, Rich Interaction, Rich Relationships
>March 22-26, 2007, Las Vegas, NV
>www.iasummit.org
>-----
>When replying, please *trim your post* as much as possible.
>*Plain text, please; NO Attachments
>
>Searchable Archive at http://www.info-arch.org/lists/sigia-l/
>
>
>
>
>________________________________________
>Sigia-l mailing list -- post to: Sigia-l at asis.org
>Changes to subscription: http://mail.asis.org/mailman/listinfo/sigia-l


          ARTHUR FINK Consulting  arthur at arthurfink.com
          ---------------------------------------------
           Ten New Island Avenue  Listening to users
       Peaks Island, Maine 04108  Designing for people
              www.ArthurFink.com  User interfaces that work
207.766.5722  cell 207.615.5722  Progress training + consulting