Human Factors

Lecture notes : Human Factors

(based on Shneiderman chapter 1)

[ lecture notes | CSC 397 | Pete Sanderson | Computer Science | SMSU ]

Five measurable human factors central to UI evaluation (conflicting):

1. Time to learn
2. speed of performance (define and time benchmark tasks)
3. rate of user errors
4. retention of knowledge over time
5. subjective satisfaction

Importance of each is determined by nature of system, e.g.
1. life-critical systems
2. industrial and commercial systems
3. office, home, and entertainment applications
4. exploratory, creative (support creative activity), and cooperative systems

In-class 10-minute brainstorming exercise. Class divided into two groups. Each group's task and results are given here. Each group has designated scribe.

Support the statement: Human-Human Interaction is a good model for Human-Computer Interaction.

1. verbal interaction - can use voice commands for collect calls for instance
2. more personal
3. less intimidating
4. good model for entertainment and help systems
5. underlining for spell and grammar checking
6. model for expert systems and similar system designs
7. a help system prompts user to send email request for help
8. the mouse as a result of HHI -- pointing
9. system can help user by suggesting alternate ways to accomplish tasks
10. good for inexperienced users

Support the statement: Human-Human Interaction is a bad model for Human-Computer Interaction.
1. computer is not dynamically responsive
2. too personal, people do not expect computer to act like human
3. user wants to be in control
4. too complicated for computer to act like human
5. room for error with person. Not for computer unless user makes a mistake.
6. we don't need to create more people

What can you conclude?

Shneiderman says that in general, HHI is a poor model for HCI!