[Sigia-l] Design by testing

[infoarchitect at ourbrisbane.com]

Ziya said:
> Oh, let's ask a few users what they think of leaving port 135 open. Why?
> Well, we don't have time to explain all this port stuff and there's really
> no earthly reason to, but what the heck. Oh, let's get a 'usability
> engineer' to see if users like the notion of becoming infected even without
> any action by the user, such as clicking on an attachment. Or if it's a good
> idea to automatically preview/render stuff on our email app, also known as a
> virus delivery device. Yeah, let's let the user make these critical
> architectural decisions for us, it's their workflow convenience we're
> talking about, isn't it?
> 
> Yep, "trustworthy computing," innovation by design, design by testing, users
> know best, ask them, compile and ship, that's the ticket.

Hmmm...  Has Ziya been 'infected' with a virus (or perhaps just 'infected' by people whingeing 
about picking up one of a plethora of virii) of late? ;)

I must say that I do agree with you on the point of 'design by testing'.  However, I don't 
discount user-centred design.  Usability testing per se is meant to verify designs - not 
create them (that's why they call it 'testing'). 

Oh, and I don't believe that blaming poor ol' Usability Engineers is getting to the heart of 
the problem pointed out in this article...  Yes, purely subjective means (questionnaires, 
focus groups, etc) are not enough to base a design upon, but a marketing-driven company "Oh, 
hello Microsoft" might tell you different (as opposed to having them "Think different", 
producing superior products that flounder in the market).  I think that you might find that in 
Microsoft, the Usability Engineers are meant to merely be the pawns that test the usability of 
products that Marketing have dreamed up.  Then again, I could be wrong.

I've said it before, and I'll say it again - people don't know what they really want or need - 
but if we analyse them, we can make informed decisions as to how to help them achieve their 
goals.  It's up to us as designers to create 'informed designs', drawing on a number of 
skillsets (ethnography, anthropology, information management, system architecture, etc), 
and/or employing a number of techniques (contextual inquiry, heuristics, task analysis, etc).  
The resulting hypotheses (designs) should be tested (for usability, system integrity, code 
validity, etc) to verify that what is produced is robust, efficient and helps users achieve 
their goals in a <insert adjective depending on context: meaningful, productive, efficient, 
pleasurable, etc> manner.

As an aside:  I recently read a paper by Vincente (1999) that outlined how the introduction of 
desktop computers to office environments costs billions each year, yet has produced a 
consistently declining ROI.  Productivity started falling with the introduction of these 
confusing machines (that constantly had problems - or were fun to 'customise') that required 
support on all fronts ("I bet Geoff knows how to format this spreadsheet better!", "Stupid 
machine!  Stop crashing when I do that", "I can't get this file to open!", "Uh-oh, I just 
deleted that whole directory", etc).  

Funnily enough, productivity faced an even steeper decline with the advent of email and the 
Internet... :P  

Speaking of which, I'd better get back to work... ;)

Best regards,

Ash Donaldson
User Experience Designer