CMPUT 301: Lecture 27 Help and Documentation Martin Jagersand Department of Computing Science University of Alberta.

CMPUT 301: Lecture 27Help and Documentation

Martin Jagersand

Department of Computing ScienceUniversity of Alberta

2

User SupportCommon Fallacies

• “It’s so easy to use, you don’t need help.”– But GUI systems can also be complex

• Read the manual, it’s all in the manual– used by gurus to brush off questions they

consider trivial

• Read the source!– used when the manual doesn’t help

3

User SupportOther fallacies and problems

• Dumb warnings:– air conditioner

– Avoid dropping air conditioners out of windows.

– blow dryer– Do not use while sleeping.

– iron– Never iron clothes on the body.

– vacuum cleaner– Do not pick up anything that is burning.

4

User Support

• Dumb warnings:– Batman costume

– Cape does not enable user to fly.

5

User Support

• Main types of assistance:– quick reference

– mostly used as a reminder

– task-specific help– for achieving a task

– full explanation– for better understanding of the system

– tutorial– step-by-step example for new users

6

User Support

• Help– problem-oriented and specific– e.g., task-based instructions

• Documentation– system-oriented and generic– e.g., user interface snapshot

7

User Support

• Online support:– get help from the computer, assuming it is

running …– Command line text based: unix man, info,

matlab help.– Interrupt current command line interaction, (unless

starts in separate window)

– GUI based – Displayed concurrently with task.

8

Online Support

• Requirements:– availability

– access help at any time, at the same time

– accuracy and completeness– help should correctly cover the whole system

– consistency– different parts and types of help should be consistent with

each other on content, terminology, and style

9

Online Support

• Requirements:– robustness

– the help system should itself be reliable

– flexibility– present multiple levels of help

– unobtrusiveness– avoid intruding on the user

– MS paperclip?!?

10

Online Support

• Command assistance:– provide help at the command level– sometimes terse and obscure– assumes user knows what to look for– e.g., Unix man

11

Online Support

• Command prompts:– provide guidance on correct usage

12

Online Support

• Context-sensitive help:– present help that depends on the context instead

of only one entry point(e.g., specific user, system state, step in a task, object being pointed at, etc.)

13

Online Support

• Online tutorials:– allow the user to learn the basics of a system

within a special environment– e.g., system with some blocked functionality to

ease risk-free exploration

14

Online Support

• Online documentation:– put existing paper

documentation available on the computer

– could use hypertext techniques

15

Online Support

• Movies:

QuickTime™ and aVideo decompressor

are needed to see this picture.

16

Online Support

• Office assistants:

17

Adaptive Help Systems

• What:– adapt the help system to the user

– similar to general adaptive user interfaces

– monitor user activity– e.g., preferences, mistakes, etc.

– build model of user

18

Adaptive Help Systems

• Issues:– easier said than done– should the system actively initiate help?– will adaptivity confuse the user?– what are the knowledge requirements?

19

Adaptive Help Systems

• User modeling:– every interactive system incorporates some

model of the intended user implicitly within its design

20

Adaptive Help Systems

• User modeling:– in some systems, the user can provide a model

around which the system will be configured– requires experience to customize the system

appropriately– e.g., Unix .profile and .cshrc

21

Adaptive Help Systems

• User modeling:– or automatically update a model of the user,

starting from a default model– Typically: Prompt for choices once (e.g. which

printer) then remember it.– Advanced: Associative content retrieval (MIT

project, EMACS demo available)

22

Adaptive Help Systems

• Quantification:– define a number of levels of user expertise– place user on a level and continuously measure

expertise– expertise is scored based on the weights of

various actions– move user between levels at a certain threshold– adapt the help system accordingly

23

Adaptive Help Systems

• Issues:– initiative

– mix of user and system control

– system suggests hints about unused functionality

– user requests help as usual

24

Adaptive Help Systems

• Issues:– effect

– how will the user model be used?

– to offer different help for novices and experts, may not need to record detailed sequences of user interaction

25

Adaptive Help Systems

• Issues:– scope

– application wide or system wide

26

User Support Systems

• Guidelines:– help should be designed as an integral part of the

system, not “bolted on”– usability principles also apply to help systems– help systems also need to be evaluated

27

User Support Systems

• Presentation issues:– how is help requested?

– e.g., command, button, toggle, separate program

– how is help displayed?– e.g., new window, integrated frame, balloons

28

User Support Systems

• Effective presentation:– clear, instructional language– consistent terminology– describe how to use the system

29

End

• What did I learn today?

• What questions do I still have?