[Sigia-l] validating an IA?

[Eric Scheid]

On 27/3/06 7:29 PM, "Listera" <listera at rcn.com> wrote:

> so speaking generically now, clients usually ask for "validation" for three
> reasons:
> 
> 1. They don't understand the deliverable
> 2. They don't trust themselves
> 3. They don't trust the producer
> 
> It's imperative to figure out which case applies.

They can ask for other reasons too:

4. the deliverable didn't include the supporting research
5. time has moved on, is the old IA still a good fit?
6. the field of IA is still nascent, best to double check
7. independent testing works better than self-testing

Nothing to get defensive about.

> Public organizations mandating "validation" is somewhat of a special case,
> but here in this instance, they are singling out a specific person/work,
> which obviously is the problematic part.

The two recent instances I encountered of an organisation asking for
validation of IA work were for cases where the only work done was the IA
work itself. Bit hard to test the graphic design when no graphic design was
done (ditto usability, business analysis, etc).

As to the validity of validation at all ... how come the same moral outrage
isn't being raised over all the usability testing of other design work? ;-)
Or is there some kind of difference between usability testing of interactive
design and validation testing of IA?

e.