Create with Context has conducted an extensive usability test of how people use the iPhone. It makes for very interesting reading.
Also, download the PDF
Writer, Project Manager and Interaction Designer
follow me on Facebook, Twitter, Flickr, or visit my website
Jonathan, you are spot on about "repetitive use". They really cannot conduct usability studies like this.
One example: When I made recently presented the iPhone app I made for the 5 clothing companies. One of the questions I got was "how do people know that can swipe their fingers to move from one image to the next"
My answer was "if they got an iPhone, they just know".
Most people who tries the iPhone seems a little confused at first, it is a different way to interact. But 2 minutes later the confusion has been replaced with "ohhh... that actually works quite well".
What I thought was interesting about the above study is that it illustrates the initial reactions that people have. The "out-of-the-box" experience.
And swiping their fingers is (probably) a lot easier than repetitively acquiring/missing/slipping off an "intuitive" on-screen control that says "Previous/Next." But really - use of the word "intuitive" should be banned from polite conversation about UI design.
Personally, I think the "out-of-the-box" experience of many products (in which I include the Mac) has been responsible for perpetuating the awful usability of things like phones. The iPhone is better than most, but in the mobile handset industry, if the unit looks cool and the home screen looks funky, the marketroids assume that's all that's needed to make it a great product. Real usability hardly comes into it.
Here's a fact to ponder about the state of mobile UI design right now: Despite mobile handset usage being more prevalent than any other digital device in the world, Jan Chipchase of Nokia is only notable UI designer in the entire industry.
Hello Jonathan and Thomas - I am the one responsible for the iPhone report, and read your comments with interest.
Our goal with the research was twofold: understand how interactions on the iPhone influence emerging interactions on other platforms, and understand how people 'natively' interact with the platform prior to becoming accustomed to its quirks and foibles.
With respect to the second goal, we find that this type of research helps iPhone developers (and, by extension, developers for other platforms) understand what interactions will be intuitive for their users - even if these interactions are not directly in conformity with the official design guidelines. Better said, if people get used to quirks, these may become a fundamental part of the interaction metaphor for the platform.
Curious to hear your further thoughts.
Baekdal.com is (nearly) everywhere. Check out our social profiles:
Published: November 19, 2008 
FaceBook
Twitter
LinkedIn
YouTube
FriendFeed
Delicious
Jonathan
What a completely pointless bit of research.
Firstly, "intuitive" interfaces are a MYTH. You cannot sit down in front of any UI of even moderate complexity and just expect to start accessing its functions without thinking hard about it and making a lot of mistakes. Anyone who says otherwise either works in marketing or is an idiot. In fact, they're probably both.
Secondly, the iPhone is designed as a tool of *repetitive* use. Conducting a test like the ones these people have is like sitting somebody in Formula One who has only ever driven a tractor, and asking them to drive around in down-town New York.
The fact is that many things that appear less usable at first (like keystrokes for instance) are in fact highly efficient once learnt. Anyone who has ever been taught to play a musical instrument, golf, or any number of skilled undertakings, will tell you that. The same is true of complex machinery like the iPhone.
So, to determine the usability of the iPhone over time would require a completely different test. Too bad these idiots conducted the wrong one. Still, at least we all know which consultancy NOT to hire now.