CMSC434: Introduction to Human-Computer Interaction

Sample solutions for Exam #1

1. D (2 points)

2. E (2 points)

3. B (2 points)

4. B (2 points)

a) (9 points, 2 points each for each pro and each con. 1 point for comparing them to each other) sample answers
Controlled experiments have at least two treatments and seek to show statistically significant differences; usability tests are designed to find flaws in users interfaces. Both strategies use a carefully prepared set of tasks, but usability tests have fewer participants (maybe as few as three), and their outcome is a report with recommended changes, as opposed to validation or rejection of a hypothesis.
Controlled Experiment:
(pro) Would be able to test whether or not new interface allowed users to find prescription drugs at local pharmacies faster, once a new interface was implemented;
(con) Would only be able to test hypotheses using measurable and controlled variables. Would need to test on many people in order to produce statistically significant results. Could not test to see what the users thought of the interface or how much they enjoyed using it.
Usability Test:
(pro)Could start testing with current web page on targeted users in order to obtain specific recommendations for changes. Feedback to guide work. Need fewer participants.
(con)Setting up a usability laboratory. May be hard to find willing sample users to devote enough time for the usability testing. It emphasizes first time usage, so users may be better able to use the web page with practice. No easy way to test that. There are a limited amount of tasks that can be tested during a usability test, but this is a “simple” web page, so there may not be many tasks.

b) (18 points)
REQUIRED
Testing method – 3 points
Testing tasks – 3 points
User Identification – 3 points
Timetable/phase structure – 3 points
Note taking details – 3 points
OPTIONS – 3 points
Room set up details
More testing method details
Pilot study/dry run/revision of TEST
Consent form
Research to understand elderly

c) (6 points: 3 points for description, 3 points for appropriateness)
GOMS: goals, operators (tools/objects the user uses in order to achieve goal), methods (the way in which the user uses the tools/objects), and selection rules (control structures for choosing between several methods available for accomplishing a goal
GOMS works nicely for describing steps in decision making while carrying out interaction tasks, such as text editing in a manuscript
GOMS is not appropriate because it emphasizes expert users doing routine tasks, and it predicts time of performance. For this case, the users are not experts, are not in a rush, and are not doing routine clicking, mousing, etc.

d) Grading: Any reasonable answer would receive credit (2 points each way described)
Some sample answers:
Use a large font
provide a text equivalent for every nontext element
Avoid relying on short term memory
Rely on recognition, not recall
Minimize the amount of typing needed/use fill in where ever possible

6. (6 points)

With a controlled study, the outcome produces (hopefully) significant statistical differences, and the people look at one or two differences. With a usability test, you do not always know what the outcome will produce. You may not have quantitative results. You only see if it is usable and get tips for things that require further looking into.

7. (1 point for each E, 1 point for each correct description, and 1 point for each correct application, 12 total)

The four Es provide a basis for creating a living document and a lively process:

Education: Users need training and a chance to discuss the guidelines. Train web designers how to use the guidelines for creating web pages for the blind.

Enforcement: A timely and clear process is necessary to verify that an interface adheres to the guidelines. Create a system that will or body of people who will check to make sure web pages created adhere to blind guidelines—or something to this effect.

Exemption: When creative ideas or new technologies are used, a rapid process for gaining exemption is needed. Some new technologies may be exempt from adhering to the blind guidelines for web pages.

Enhancement: A predictable process for review, possibly annually, will help keep the guidelines up-to-date. Every year, review the guidelines for blind users of the web to bring it up to date with new technologies

8. (1 point for each attribute, 2 points for each correct example, 9 total)

Advantages of direct manipulation, summarized in three principles:

Continuous representations of the objects and actions of interest with meaningful visual metaphors
Physical actions or presses of labeled buttons, instead of complex syntax.
Rapid, incremental, reversible actions whose effects on the objects of interest are visible immediately.

Other beneficial attributes:

• Novices learn quickly

• Experts work rapidly

• Intermittent users can retain concepts

• Error messages are rarely needed

• Users see if their actions are furthering their goals; if actions counterproductive, they can simply change the direction of their activity.

• Users experience less anxiety because the interface is comprehensible and because actions can be reversed easily

• Users gain confidence and mastery because they are the initiators of action, the feel in control, and they can predict the interface’s responses.

a) Grading: Any reasonable answer would receive credit (2 points each mistake, 6 total)
Several of the many possible things wrong:
consistency of the font type/size/color
visibility—the title is light grey on light grey
alignment—of the buttons, the labels, the text fields, the group boxes…
no help button/link
buttons are not consistent
each label for a textbox has different background color and font type
are you supposed to put in last name or full name?
no avoidance of user error—in the date field, reason code

b) These are the most obvious rules, others as long as they make sense(6 points):
(3 points each: 1 point for the name, 1 for the description, 1 for the reason)
(descriptions are in the book)
Strive for consistency—so many examples
Prevent errors—the date, and the name
Support internal locus of control—full name and last name labeling one text box
Reduce short term memory load—have to fill in reason code
Cater to universal usability—small fonts, no novice/expert differences, color choices

c) The Eight Golden rules: 4 points (2 points each: 1 point for name, 1 point for the description)
(descriptions are in the book)
Strive for consistency
Cater to universal usability
Offer informative feedback
Design dialogs to yield closure
Prevent Errors
Support internal locus of control
Reduce short-term memory load

10.(10 points total, 2 points each; 1 point name, 1 point explain)

a) There was no a.

b) Layer 1/Windowing System layer Allows you to build all of the application, not for a specific domain

c) Layer 4/Application layer or 3/Application Framework/Specialized Language layer Both layers 3 and 4 have a short build time, but if you also need to learn the tool, then layer 3 has the shortest learning time

d) Layer 3/Application Framework/Specialized Language layer Has the shortest learning time of the layers, takes only days.

e) Layer 4/Application layer Best one of the four for extensibility and modularity

f) Layer 1/Windowing System layer Of the four, this one communicates the best with other subsystems across the domain of the tool.