View this PageEdit this PageAttachments to this PageHistory of this PageHomeRecent ChangesSearch the SwikiHelp Guide
Hotspots: Admin Pages | Turn-in Site |
Current Links: Case Final Project Summer 2007

Sum2003 Midterm Review: User Interface

Considering we cannot use a book on the test, how would we go about answering number 5?

If this question is on the midterm it would have a picture of the interface with it. Barbara Ericson

1. You would want to know if the children could read. You would want lots of pictures and animations. Do not use large words and have some kind of speaker giving the instructions.

2. You would need large text so that they could read. you wouldn't want to have too many scroll bars or mouse movements (quick reactions are not good for old people).

3. Before: Make sure information is visable and is in the language of the reader and not the programmer. Help is easily accessible. Error messages make sense for the user (not error 3003 occured!).

After: Watch the user interact with software (don't help them). Ask why they did something. Interview them to see if you can add any functionality or make things more convenient.

4. A cognitive walkthrough is good before you talk to the user. You want to imagine you are the user and make sure that the system makes sense. Also make sure you've fufilled all the requirements.

5. The time is too small and the buttons are too big. Since the buttons aren't the focus of the clock, they should not be the focus of the display. Also, once you've clicked a button, you do not have any way of knowing if it worked until the time changes (the button doesn't change).

6. The waterfall design is not a good way to design a UI. The iterative design and prototyping is much better because the user is constantly involved.

Good job Laren and Dennis. I agree that iterative development is better than waterfall but I am looking for a bit more about how we create the user interface (hint, users, tasks, etc)


Cognitive walkthroughs were first proposed by the quartet of Polson, Lewis, Rieman, and Wharton in their seminal paper 'Cognitive walkthroughs: A method for theory-based evaluation of user interfaces', in an effort to improve upon informal walkthrough evaluation techniques through the introduction of psychological theory [Human-Computer Interaction; Dix, Finlay, Abowd and Beale; 409]. In a follow-up paper, 'The cognitive walkthrough: A practitioner’s guide', Polson, et al revised their idea of the cognitive walkthrough, with an eye toward making it more accessible to designers.

Simply put, a cognitive walkthrough involves the identification of a sequence of actions that describe the accomplishment of a task in a system, and the stepping through of that sequence by usability experts looking for usability problems. For each step in the sequence the evaluator attempts to develop a believablility story by answering the following four questions:

1. Will the users be trying to produce whatever effect the action has?
2. Will users be able to notice that the correct action is available?
3. Once users find the correct action at the interface, will they know that it is the right one for the effect they are trying to produce?
4. After the action is taken, will users understand the feedback they get?

Michael Mitchell

Nice job Michael Barbara Ericson
1. When developing a UI for K-12, you have to consider the fact that most little kids do not know how to read and if they do know, most of them would like seeing a demo better than having to read. You can also include animations, talking instructions, fun images, and less text.

2. When developing a UI for senior citizens, you have to consider their eyesight and reaction time. Most older people have trouble reading small font, so you want to have bigger font than usual. Also, their reaction time is slower so you might not want to have a lot of mouse usage or typing. You have to be careful about what vocabulary to use and you have to be aware of their knowledge.

3. To test the UI before the users are involved you can have a cognitive walkthrough, going through all the steps as you are completing the task. You can also check the user interface, check the guidelines, make sure error messages are meaningful, and that knowledge is visible. To test UI after the users are involved, you can develop a questionnaire, observe the user while they complete the task, tell them to say outloud what they are doing as they are completing it, and try to figure out why they do certain things.

4. Cognitive walkthrough is imagining you are the user who is walking through the interface performing the task. You are trying to figure out if the system makes sense. It is useful to test the UI before the users are involved.

I dont' think this is quite right. As the designer of the UI you're probably too close to the problem to objectively evaluate it. Someone else, who is an expert on usability, should perform the walkthrough using the formal process listed above. I think without the development of the believability story by an independent party it's not a cognitive walkthrough but some folk hci derivative. Similarly, I think the slides are misleading with respect to the heuristic evaluation. Central to this method as well is that someone else perform the evaluation. In this case preferrabbly 5 hci experts should independently evaluate the interface (Jakob Nielson, one of the creators of the heuristic evaluation estimates 5 evaluators will find 75% of the usablilty problems, with diminishing returns beyond 5). Michael Mitchell

5. Most of the attention goes to the buttons which are changing the clock, which is not something we do most often with a clock. Most of the time we want to see what time it is or set the alarm and the font on this UI is too small. Also, you're not completely sure whether you've clicked on the buttons to change the clock because it doesn't make any type of sound or change its display.

6. To develop a good UI, you have to realize that you are not the user and that you have to know your users. You have to know their ability, knowledge, reason for using your software, etc. Make sure knowledge is visible.
~Sabina Karkin
Your answer to 6 is more what I was looking for Sabina, except I would want to see something about user tasks. Your other answers are good. Barbara Ericson

Link to this Page