Feb 25th, 2003
Your best guess isn't good enough

If it were interfaces wouldn't be so bad! Some designers care and try
Think in terms of user and tasks give it your best shot evaluate and try
again
Willingness to do so = maturity (like debugging)

The user is always right (comes from customer is always right)

They aren't stupid. IQist thinks users are stupid. (like racist, sexist)
Some social problems to solve in the world correlates with people who
are iliterate and uneducated.  

They aren't lazy. 

By definition, if they have trouble with the interface, it's the fault
of the interface (and it's designer)

Caveat: Assuming the users are motivated towards a task and are focus on
it. If people are not interested in trying, then it is a big problem. 

The user is not always right

Inertia - resistance to change
Better at finding problems than suggesting fixes
Often suggest many specific tweaks with no global consistency
React to artifact, not concept or generalization
Users are highly variable. 

So, when do you listen to them?

You always listen - grist for your mill
When do you act? (acting/changing is expensive)
-	when user testing shows a flaw
-	when users react strongly positively or negatively
-	when > 1 person says it

Users are not Designers

BTW, you guys are designers! Caring about how things ought to be.

User customization is not the answer.
Novices are afraid to customize (and rightly so!)
The power of the default <nobody changed the default even when it's easy
to do>
-	"suit grey" (and window size -> DOS) 
-	X and Uwin (tile the screen surfaces)
Customization -> differences -> NO informal help
-	You lose informal help
-	You put users in space, you are requiring them to customize.
They can wind themselves into awkward situation where things don't work
well together any more.
-	Users are afraid to customize

Bad choices: 
	Command abbreviations and error rates!

Designers are not users

Similarity: Designers are human. They use the technology.

Differences: General experience, enthusiasm (we're geeks!), mental model
of implementation

You can't "unlearn" something, and it taints your view of the system.
(this is why I prefer UTAs to TAs) Empathy is so important for TAs. Use
people who are closer to the target population so that you get the gut
reaction.

We take Dad to store and show him Macintosh. He touches the screen
(kiosk). Takes the mouse, points at screen and presses the button
(remote control). Scottie talks into the mouse in Star Trak.  

At the end of the day, after designing anything, you user test

Vice Presidents are Not users

Neither are CEOs or professors
They hold great influence and sometimes meddle
They were promoted for their managerial skills and not their design
skills
But if they care, the battle's half won, so nurture them! (and explain
about testing!)

If you cannot make an opinion get heard, give facts behind the opinion.

Less is More

Lazy designers and cowards throw in the kitchen sink.
Think of the interface as a backpack on a 10-mile hike - cognitive load
Remove components by 
-	ruthless removal of power
-	good design
Seldom used components are an opportunity cost on user time and
cognitive load
 
Details Matter

The devil is in the details
Weak link in a chain
Single work choice in a manual
e.g. Do not use "Last"
"in the last step.."
Previous or final? What is the meaning of Last?

Help Doesn't
Layers of help doesn't fix underlying bad design structure
Online help is like fraternities - those who really need it can't get in

Context-sensitive and balloon help address this.

The people who are in the most trouble just cannot be helped. 

Help and documentation will not save bad design
"Push" on door with handle sign.
