Homework 3

Homework 9: Model-based Evaluation

Due Thursday, April 3

Although usability evaluation is sometimes purely empirical (counting errors or timing task performance), it can be a more powerful tool when it is combined with models of various sorts. For example, user experiences can be categorized and assessed using the claims that originally guided the design and development work. In this homework, you will develop a simple GOMS model for the tasks of "resetting the time to one-hour earlier than the current setting" and "verifying and activating the alarm".

Here is a GOMS model for the task of “verifying and activating the alarm” for my own clock radio. (Note that I have assumed that the time is set correctly; a more complete model would include an optional embedded method for setting the time as well.)

Goal: Turn on alarm
      (sub)Goal: Verify current alarm time
            Operator: Visually scan radio top for ‘Alarm’ button
            Operator: Visually scan radio top for ‘Set’ button     (Verify method)
            Operator: Move thumb and finger to 'Alarm' and 'Set' buttons
            Operator: Simultaneously depress ‘Alarm’ and ‘Set’ buttons
            Operator: Verify clock display to read ‘5:45am’
      (sub)Goal: Set alarm to be on (can be implemented with two different methods)
            Select Goal: Push and correct
                  Operator: Locate slider by touch
                  Operator: Push left until stop
                  Operator: Pull back gently one ‘click’
            Select Goal: View and adjust
                  Operator: Scan radio top for ‘Mode’ slider
                  Operator: Move slider to ‘Radio on’ position
                  Operator: Verify ‘Alarm on’ dot visible in clock display

First, develop two GOMS models at roughly this level of detail for your own clock radio, alarm clock, digital watch, etc. for each of the two tasks "resetting the time to one-hour earlier than the current setting" and "verifying and activating the alarm". If you do not own and cannot get access to any devices in this category, you can make up an analogous exercise using a similar device, such as a VCR, microwave oven, ATM, and so on. Also, note that I am using a sort of pseudo-formal GOMS notation; you do not have to be too concerned about the formalism as such.

Second, discuss the assertion that GOMS models like the ones you have developed could provide the basis for analytic and mediated evaluation of user interactions. For example, do simpler user interfaces for the same task would have simpler GOMS models? Could GOMS models plus empirical data on task performance times be used to develop standard estimates for routine task components like scaning for/choosing a button, pushing a button, verifying a setting, and so forth?

What to turn in:

Your paper should include your two GOMS models, a brief description of the models, and your analysis of the assertion that models like this could provide the basis for analytic and mediated evaluation of user interactions.

Format your paper in 12-point Times or Times New Roman font, single-spaced. Write no more than two pages (including any image file you embed in the document), printed duplex (that is, hand in a single sheet of paper). Your name and your student number (SS#) should appear in the first two lines at the top-right of the first page in bold; the heading "Homework 9: Model-based Evaluation" should appear at the top-left of page 1 in bold.