Changing Horses-October to December 1998

The first evaluations took place during October and December. We wanted to get the tests started as soon as possible, so we took some shortcuts. Instead of using a large, carefully selected user sample, we ran the tests with just two of our colleagues. Also, the weight of our prototype did not allow handheld use, which limited the validity of the results. However, by making these compromises we were able to get some feedback at a very early stage of development. It turned out to be extremely useful.

At this stage the team was primarily interested in how users learn to touch-type. Therefore, instead of running several short tests, we ran a longitudinal experiment with fewer subjects. Subjects used the prototype one-row keyboard for a full 4 weeks, writing about 20 short texts each day. In total, they wrote over 24,000 characters during the experiment (which is very close to the amount of characters in this chapter).

Both subjects were regular mobile phone keypad users, sending tens of text messages per day. They had also been experimenting with T9, but didn't use it on a daily basis. Both used a PC keyboard daily and were relatively fast typists, using all 10 fingers for typing. However, neither could do real touch-typing. The subjects were instructed to 'touch-type' throughout the test by using four fingers of both hands, each finger occupying one key on the keyboard. space and clear were operated by moving the left little finger off the abc key. Thumbs were held idle.

During the evaluation, it became obvious that touch-typing with the one-row keyboard was not as easy to learn as we had hoped and expected. Typing with one-row was very slow in the beginning, since learning the new key locations took a while. For quite a long time subjects had to visually 'hunt and peck' each key. The final typing speed was also lower than we anticipated; at the end of the test both subjects had attained less than half of their PC keyboard speed.

The most remarkable finding, however, was the vulnerability of the touch-typing ability. All went well enough as long as the subject was well rested, alert, and typing in an undisturbed environment. However, at the most minor disturbing factor, performance collapsed. Both of our subjects explicitly mentioned that typing was burdensome, and that they had to concentrate on it all the time to get the job done with any accuracy. The human factors expert in our group had a name for the burden; he explained cognitive load and working memory capacity. It was obvious that touch-typing took up the subjects' entire working memory capacity-or at least most of it, and then failed when other loading factors appeared. This was a huge problem, even a risk factor in the mobile context. We obviously did not want to see people running into accidents because they were using all their cognitive capacities to operate our fine new keyboard.

This result led the team to an obvious, but painful conclusion: get rid of touch-typing! The project changed directions, and went after what we called 'four-finger typing' instead. It is similar to what many untrained typists do on their PC keyboards; instead of real touch-typing, they use a couple of fingers-two, four, or more depending on their expertise-in a relatively freeform manner. Four-finger typing seemed quite easy to learn at least for users with one-row experience. Our two subjects were able to reach their touch-typing speed in a matter of minutes.

Getting rid of touch-typing gave us one additional advantage. We could further reduce key width without compromising usability too much, as the fingers no longer had to fit side by side on the keys. Once again, the hardware designer went out looking for different key samples and key widths. This time we wanted to make the prototype more realistic, and issues like key 'travel' and tactile feel were considered. The result was a prototype realized with small switches with exceptionally low travel and a very snappy tactile feel. The key spacing was reduced to 12.5 mm (1⁄2 in), a size that should be comfortable for four-finger typing. The full width of the keyboard was now reduced from 160 mm (61⁄4 in) to 134 mm (51⁄4 in) (Figure 7.4).

click to expand
Figure 7.4: The second one-row prototype. We give up on the idea of touch-typing and reduce the key pitch to 12.5 mm. The total width is now 134 mm.

The whole device would be somewhat bigger than a plain mobile phone, but still be small enough to be carried around in a pocket. The team was now ready for the second user test.



Mobile Usability(c) How Nokia Changed the Face of the Mobile Phone
Mobile Usability: How Nokia Changed the Face of the Mobile Phone
ISBN: 0071385142
EAN: 2147483647
Year: 2005
Pages: 142

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net