Kevin Schofield's Weblog Musings on life, kids, work, the Internet, Microsoft, politics, orcas, etc.

Home Microsoft Research Danceworks Studio

Thursday, September 22, 2005 A Medical Breakthrough? This could be the start of a huge new direction in medicine -- the ability to regrow limbs and organs. It's early work, and probably a decade or two away from practical use, but very promising. 9:18:25 PM    comment [] One More Cool Thing About Microsoft Wow -- I haven't been in a really meaty HCI conversation in a while, and I didn't realize how much I missed it until tonight. Here's a little known fact, but probbaly won't surprise people once they ponder it for a moment: Microsoft has a central archive of reports from usability tests and user studies on ts products. The reports date all the way back to 1989. I just spent half an hour rooting around in it. The archive is an amazing historical treasure chest; among other things, it contains the usability test reports for Windows going back to 1989, and for Office back to around 1994. It also contains the usability test reports for Microsoft Bob. (Yes, it was usability tested. It did really well, too. Which is one reason why it was released. It failed for completely different reasons, a topic for another time) Now, anyone who has worked on a user interface at Microsoft can tell you that there is an amazing machine inside the company that does user studies. Our user study specialists are very detail-oriented and very, very disciplined. They can also be quite creative, and particularly in the last few years as they've dug deeper and deeper into the mysteries of how our customers use our products -- or would LIKE to use them. They have done thousands of tests over the years, and involved tens of thousands of users. The scale of the undertaking, across our entire product line, is simply staggering. A number of years back, I floated the idea internally that we should release into the public domain the usability tests and user studies that are old enough to no longer be of competitive or strategic value. My thinking was that as historical documents on the process that went into the design of such high-impact products, they would be of significant value to the academic community. And frankly speaking, it would probably get us some well-deserved credit for a lot of hard work that has never really been exposed to a public that likes to complain about our "feature-bloated" products and their personal pet peeve. In the end, I couldn't get approval to do it, for one excellent, very simple and straightforward reason: because we promised we wouldn't. As a condition of people coming in and consenting to trying out some software and letting us observe, record the session and analyze the results, we promise to keep what happens in the session confidential. It's an important guarantee, because that's the only way for us to get honest, real results. People get embarrassed when they make mistakes in a public forum; if we don't go overboard to reassure people that the session will not be exposed to the public, then we have to expect that they will participate less than fully. It's absolutely sensible, it's the Right Thing for the people who help us with our user studies, and when it was explained to me, I stopped trying to get the usability reports released. But hey... if you come work for Microsoft as a full time employee, you can peruse the archive yourself! 8:01:50 PM    comment [] Menus and Dialogs and Office, Oh My! OK, I've found myself in the midst of what has become a really interesting discussion over here. I started writing a response in the "comments" section there, and it got very long, so I'm moving it over here. I encourage you to go read the thread from the beginning, especially if you are into user interface design. This conversation started as a critique of the Office 12 user interface, which has been drastically reworked in this release. Soeren, you have many interesting ideas. You could argue that in the earlier days of Office, with OLE used for many embedded objects, they were sort of doing exactly what you said: have a separate application for various things. Context menus (i.e. right-clicking) is also the same general notion. The Office team has experiemented with all sorts of twists on this over the years. Early on in any product development cycle, they prototype several different approaches and test them out. They especially do this when they are biting off a major re-vamps of the UI -- which fortunately doesn't happen very often. Let me throw something else out here for you to chew on: an alternate explanation to why cascading menus are only effective for 2 levels of depth. It turns out that studies show a strikingly similar effect for file/folder hierarchies: most people,when left to their own devices, never go beyond two levels of depth of folder hierarchies. And it correlates very highly with math/science/technical education: the more your education in those areas, the more likely you are to have deeper folder hierarchies.  Here's the theory: it has to do with Piaget's stage theories of cognitive development. The top two stages of cognitive development in his hierarchy are "concrete operational" and "formal (or abstract) operational". At the concrete operational level, you can manipulate in your head ideas andobjects in two dimensions. At the formal operational level, you can do it in three dimensions (and sometimes more). Piaget's experiments found that most people never got to the formal operational level in their lifetime, but he also found a high correlation with math/engineering/science education. Sound familiar? Now, I'm not implying a particular cause/effect relationship; there's a huge debate about whether you can actually teach formal operational abilities, or whether they're innate and simply emerge at a certain level of physical development/age (Piaget believed the latter). But if we go back to cascading menus and folder hierarchies, it's a rather moot point, because my working theory is that each level of a menu or folder hierarchy is essentially an independent dimension in terms of how we navigate (and there are PLENTY of studies that show that people navigate both menus and folder spatially) so the ability to navigate folders or menus more than two levels deep directly relates to acquisition of formal operational abilities (i.e. can you mentally manipulate in 3+ dimensions). So while it may be marginally more difficult to navigate multiple-cascading menus with a mouse, I really believe there's something deeper going on. 7:08:57 PM    comment [] GooglePark This is just too funny. And I like that it pokes fun at everybody equally. 8:18:46 AM    comment [] © Copyright 2005 Kevin Schofield. Last update: 10/4/2005; 3:19:22 PM. September 2005 Sun Mon Tue Wed Thu Fri Sat         1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30   Aug   Oct