DTUI Booksite Assignment 4: Interface Trade-offs Analysis [Assignment 3] [Assignments Menu] [Assignment 5]

Goals:

An HCI professional would be required to know what the trade-offs are between design decisions. She or he would also be required to evaluate interfaces by doing well-designed and well-controlled experiments and by doing surveys of users.

Task:

Hand In:

1. General Comparison:

For the General Comparison, try not to include ``unusual'' commands that are rarely used. You are to compare the interfaces in general first, in HCI terms, based on the list of Interface Issues below. Watch for your own bias, and experience, as well as those of the people you talk to.

• Compare the user interfaces in general, including the speed of use and response, the choice of commands, and general quality, and other Interface Issues from the list below.

• Contrast the command ``names'' used by each system.
• Which do you find more natural? Why?
• Which types of users might prefer which system? Why?
• Are your biases showing? Confess!

2. Experiment:

The Experiment to be done is not well enough designed or controlled to be a real HCI experiment. However, it should give you some idea of this style of research.

Without considering the actual commands in the systems, design a ``representative sequence of simple tasks'' to be done on both of the systems.

For example, a reasonable, ordered subset of the following would be appropriate:

finding a file given its name,
copying a file,
deleting a file,
renaming a file,
activating an editor (Note: do not include editing commands),
deleting an old file,
getting help,
typing a file,
looking at files in a non-current directory/folder,
printing a file.

Observe one or two users of each system as they follow that sequence of tasks. Note the number of actions -- keystrokes, mouse clicks and hand movements. Make both users experts, or both novices, but don't mix them. One of the users could be you.

Calculate how many actions this sequence of tasks takes using the command languages on both systems. Obtain the count for the full command -- do not allow "standard" abbreviations. Do not allow special redefinitions of commands, or macros, but do look for the smallest number of keystrokes (i.e., use the correct commands for the job).

• Describe the procedure followed.
• Include all the results in your report.
• Evaluate your results (i.e., what do they mean?).
• Could the results be predicted from looking at the command input method?
• What are the flaws in your experiment?

3. Survey:

The Survey to be done is also not well enough designed or controlled to give real HCI results. However, it too should give you some idea of this style of research. You are to prepare some questions (about 5-10) designed to detect people's views on the strengths and weaknesses of both systems. You should ask the questions of about 5 people (preferably not just those currently in this class).

• Describe the procedure followed.
• Include in your report the summarized results of asking these questions.
• Evaluate your results (i.e., what do they mean?).
• What are the flaws in your survey?

Interface Issues to Consider:

 

Task Analysis: What tasks are supported?

Critical Steps: How easy is it to make mistakes? Where does one need to be careful? Does it report the user's errors? Are errors explained? Corrections suggested? Is it easy to recover? Can typing mistakes be fixed? Can information be recovered once deleted? Does it give warnings?

Time: Does the interface allow/encourage one to work quickly?

Display Aspects: Is information displayed clearly? Is it easy to understand?

Feedback: Does the system tell the user what is happening or has happened?

Control Aspects: Are the commands (words or icons) easy to learn? Are commands well named and mnemonic? Are icons well designed and memorable? Are things clear and logical? (e.g., order of arguments; layout) Is there a reasonable difference between the names/icons of different commands? Is the command layout clear?

Stress: Are there any stress situations?

Performance: With which is it easier to complete a task correctly? With which is it easier to complete a task quickly?

User Type: Which tends to be preferred? By what type of user? Why? Does the system cater explicitly to any particular type of user?

Quality: Is it obviously a "quality product"?

Display-Control Compatibility: Do commands ever produce displays that are incompatible with the task? For corrections, is it clear where inserts and deletes will occur?

Population Stereotypes: Do commands do what is expected of them? Do commands tend to act like those of other systems with which users are familiar? Is there a bias towards the expectations of any type of user? (e.g., naive) Are commands named the way the population would expect them to be?

Learning: How easy is it to learn? Which is easier to learn and why? How easy is it to remember? (e.g., after 6 weeks of no use)

Naturalness: Are the features of the interface "natural"? Is it consistent? Are expectations about what can be done in a certain situation usually right?

Flexibility: Can all required tasks be achieved? Does each task have a separate command? Are any commands effectively combinations of others? Can more complex commands be defined? Can commands be renamed? Is it too flexible? (i.e., lots of "baroque" capabilities which are never used and are intimidating) How does the number of commands required to achieve some set of tasks compare?

Complexity: Are there any "complex" operations? Should they be so complex?

User Assistance: Is any help available? When? Are the messages good? Does it have naive and expert modes? (e.g., verbose or not; or structured) Is any assistance provided when typing commands? (e.g., command completion)

History: Is a command history kept? Is it easy to see? Is it easy to reissue a command?

Please send comments and suggestions to the Booksite Director Last Updated: 26 November 2000