This is a WWW page originally created by students (see other credits on this page) but now copied and stored by Steve Draper as part of the CSCLN ATOM.

Lecture 11 Arenas & Gulfs

Last changed 25 Feb 1998 ............... Length about 900 words (6000 bytes).

Q&A format lecture notes for lecture 11: Arenas & Gulfs

Header details

Who maintains the web page: Roddy Neil
Who has editorial responsibility: Roddy Neil, John MacKay,Martine Petale,Claire Fullagar.
Lecture 11 (Tuesday of week 6)
Topic: Arenas & Gulfs
See the Task Analysis handout pages 4-6 .

The questions

What is an arena ?

An arena is a planning space for a user, defined by the set of objects and operations and their possible states. The theory of action describes the mental steps a user might in principle go through in operating in some particular arena. Example Say you wish to cut and paste some text.Then there is a list of arenas for this operation. select a menu and header --> Arena with issues about how selecting one item you select a menu or header. which in turn depends on the Menu operation and the Mouse operation Ō double - click Ō move up the mouse and reposition it There is a hierarchy of arena levels. When you pick a particular arena then you have gulfs between arenas. semantic ---> central arena ----> articulatory

What exactly are Gulfs ?

There are two types of gulfs that exist between arenas :

i ) Articulatory Distance This concerns trouble in executing a plan on the machine i.e you know what you have to do to the machine but it is difficult. Some examples of this are the buttons on wrist watch calculators are hard to operate and using the mouse to draw fine objects. ii ) Semantic Distance This concerns trouble in relating a plan in terms of machine commands to the terms in which a user naturally thinks about the problem domain, particularly if they already know about that domain in other contexts. Examples of this are Video recorders : They require the user to program things to tape according to the 24 hour clock.However TV listings are given as am or pm so this leads to errors. Pocket Calculator : If you wish to calculate the expression (3 + 4 + 2) / 5 then the numerator has to be worked out firstly on the calculator then you divide. Hence pocket calculators do not let you type in arithmetic calculations just like you would write them on paper.

What is Perception and Information Delivery

Interaction with computers is not just about actions, but about the information used in performing these actions.There are two kinds of information : state information, and how-to or affordance information. State information is used for verifying goal achievement, and identifying initial states on which to build plans.

Actions can in general be regarded as having two classes of effects : material effects which change the state of the program or world. Informatic effects which change what can be immediately perceived. Although some actions have only one or the other kind of effect, many have both.
Goals may be either material or informatic : users in general may be acting to achieve either a material or an informatic effect.When an action has both kinds of effect , you cannot tell the userÕs intention unless you ask them.

If the user does not have the information for their goal then they will attempt a smaller goal first.People however react to new information whether or not they were looking for it.People are also capable of noticing facts that are not currently needed but may remember them later.

Information flow is very important for users.Sources that users view frequently are

a) Memory b) Immediate Perception ( e.g. of the screen display ) c) Information Requests ( e.g. using online help, asking another user ) d) Experimenting with commands.

Three issues that lie behind this spectrum are :

a) Length of time to get the information. b) How much must you know to get the information ? c) How dangerous is it to explore ? Information types and the 3 perspectives on an arena There are basically two classes of information : state information and affordance information, corresponding to objects and operations respectively - to information about the state the system is in and information about how to change it.

i ) Corresponding to the semantic perspective is information about what context a command or operation is used in : as part of what larger plans ? E.g. the paste command is mainly used in moving text as part of a cut then paste plan.

Semantic information is often left to the user to discover : for instance if you want to move text you just have to guess that cut and paste are relevant.

ii ) Corresponding to the central perspective is affordance information about the immediate material effect of the command.

E.g. paste copies the the current contents of the clipboard if any to the current insertion point. State information for text includes font and style.The central perspective objects is usually well expressed in screen displays, and also in reference help and manuals.If a user canÕt distinguish what an object or operation does from their name, experiment will often reveal it.

iii ) Corresponding to the articulatory perspective is affordance information about how to operate a command or feature. E.g. To use a menu on the mac requires that you know how to move the mouse to the menu header, press the button, move the mouse down to the command , and release the mouse button.

Due to the fact that articulatory methods usually follow conventions, they are sometimes left wholly unexpressed in interfaces and rely on user training. An example of this is that on a keyboard there is no information to express which of the keys are shift keys.



Helpful Links to other related pages



Back to Lecture Ten

Forward to Lecture Twelve

Index

Main Page


Credits and Acknowledgements

Page designed and constructed by Roddy Neil, John MacKay,Martine Petale and Claire Fullagar.

Thanks to all those who contributed with comments / e-mail.