Top Banner
Processes of Design Fifth lecture: Evaluation 24 October 2003 William Newman
13

Processes of Design Fifth lecture: Evaluation 24 October 2003

Dec 30, 2015

Download

Documents

tarik-miranda

Processes of Design Fifth lecture: Evaluation 24 October 2003. William Newman. Evaluation: What do we Mean?. In the dictionary: Determining or estimating the value of… In interactive system design: Identifying faults Testing perfor- mance Rating quality,eg from 1 to 10. - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Processes of Design Fifth lecture: Evaluation  24 October 2003

Processes of DesignFifth lecture:

Evaluation 24 October 2003

William Newman

Page 2: Processes of Design Fifth lecture: Evaluation  24 October 2003

Evaluation:What do we Mean?

In the dictionary:Determining or estimating the value of…

In interactive system design:• Identifying faults• Testing perfor-

mance• Rating quality,eg

from 1 to 10

QuickTime™ and a TIFF (LZW) decompressor are needed to see this picture.

Page 3: Processes of Design Fifth lecture: Evaluation  24 October 2003

Evaluation in the Inner Loop

• Assessing design options• The ATM design session

D. Well, one of the things you could do is, presumably you could just whisk your card through, rather than have to put it in the slot and get it back, and that might be quicker than waiting for it to come out. And I don’t see why they don’t do that always, anyway. I suppose it’s so they can chew it up if they want to.

J. Yes.

• Evaluating by simulating.

Page 4: Processes of Design Fifth lecture: Evaluation  24 October 2003

Alternatives to simulation

• Evaluating with theory-based lawse.g. Fitts’ Law:

Tpos = K log2(A / W + 1)

• Evaluating with an informal rule or heuristic• The quest for accuracy and completeness.

Page 5: Processes of Design Fifth lecture: Evaluation  24 October 2003

Two ways to evaluate

• Empirical: conduct an experiment• Analytical: evaluate on paper• Why two ways?• Cost of simulation.

Page 6: Processes of Design Fifth lecture: Evaluation  24 October 2003

What do we simulate?

• The technology• The user• The context• The task

• Why are these familiar to us?

Page 7: Processes of Design Fifth lecture: Evaluation  24 October 2003

Evaluating in terms of requirements

• The problem statement:• Design a camera-based text-capture system

to enable students in libraries to copy text from paper to word processor in less time than by typing.

• A concise statement of requirements• As we expand requirements, we refine our

evaluations

Page 8: Processes of Design Fifth lecture: Evaluation  24 October 2003

Simulating the technology• How to avoid building the complete and final

system?• Working and non-working simulations• What’s missing from non-working sim’s?• Wizard of Oz methods.

Page 9: Processes of Design Fifth lecture: Evaluation  24 October 2003

Xerox Star

• Designed in the late 1970s• Technology performance

problems• Escalating costs• Never fully evaluated

Page 10: Processes of Design Fifth lecture: Evaluation  24 October 2003

Simulating the User

• You’re not them!• Knowing the user• Xerox Star again

Page 11: Processes of Design Fifth lecture: Evaluation  24 October 2003

Simulating the Context

• Taking account of:• Other systems (Airport control, Aegis)• Other people (GPs)• Physical environment

Page 12: Processes of Design Fifth lecture: Evaluation  24 October 2003

Simulating the Task

• Sets up the process of evaluation• Easy? We design the task• Not so easy? The user can redesign it

(the computer game)• and may never learn it.

Page 13: Processes of Design Fifth lecture: Evaluation  24 October 2003

The Limits of Evaluation