Return to HCI page

HCI test 2006 - in progress

1. pointing and fitt's law
1. a and b are experimental constants which can be roughly approximated to a=50 and b=150 for quick and dirty calculations. A represents the distance that to the object or button, and W represents the size of the object or button. A redesign based on A might be to keep the palette close to the mouse, perhaps by opening it when a secondary mouse button was hit. A redesign based on W might be to make palette near the edge of the screen so that people wouldn't overshoot the target thus making the effective size of the target larger.
2. .
1. .
2. from the perspective of fitt's law, the effective size of the menus are much larger when they are near the side of the screen because people can't overshoot. This makes it much faster for users to get to the menus.

1. User Testing

• readings not online. four problems might be: 1) taking only cs students looking for extra credit. this might bias evaluation. 2) having everyone use first microsoft excel then the other system could bias data because you've seen one system more recently. 3) telling the students that “i had been building it for 3 years” might bias them towards thinking it was good since they were motivated by extra credit and because you are also called the microsoft system the “old” system and this one seems new. 4)no comment on the statistical significance of the difference in ratings of the two systems. with only 5 students it might not be significant that the numbers are 3 vs 3.5, especially with such a coarse scale to rate them on.

1. Heuristic Evaluation
1. heuristic: user control and freedom. Problem: there is no clearly marked cancel or “emergency exit” to the function if it was clicked incorrectly. Solution: instead of a help button put a cancel button, or in addition to the help button.
2. it is difficult to satisfy the 'recognition rather than recall' heuristic on speech interfaces because speech is not visual. you don't have the various options always sitting on a screen like you do in graphical user interfaces. You can try to address this issue by having the speech interface always speak the various options that you can choose thus allowing the user to recognize functions instead of just having to remember them. the trade off here is that it can be very slow to have to listen to all of the possible options being spoken, and the options are ephemeral meaning that they don't persist after being spoken so they may have to be repeated several times.

1. design patterns
1. .
2. .
3. .
1. .
2. .

1. task analysis

1. experimental design
1. .
2. .
3. .

Return to HCI page