The Degree Navigator Nightmare: Taming An Overly Graphical Interface

1
The Degree Navigator Nightmare Taming An
Overly Graphical Interface

• Semion Bezrukov
• Alec Browne
• Kanta Jiwnani
• Donna Malayeri

2
Why this project?
3
Redesigning the Interface

• 2 Major Flaws
• Use of Inappropriate Metaphor
• Metaphor does not meet the criteria given by
  Erickson metaphors should be structured,
  relevant, representative, understandable, and
  expandable to related target domains
• Overuse and Incorrect Use of Color

4
Redesigning the Interface

• 2 Major Flaws
• Use of Inappropriate Metaphor
• Metaphor does not meet the criteria given by
  Erickson metaphors should be structured,
  relevant, representative, understandable and
  extensive
• Overuse and Incorrect Use of Color

5
Our Improved Interface
6
Background Studies

• Smilowitz
• Good metaphors improve performance
• Essential characteristic of a good metaphor is
  its mapping to the target domain
• Mayer Many programming constructs were learned
  more easily when they were presented in the
  context of a concrete metaphor.
• Foss, Rosson and Smith Subjects performed better
  while using an interface designed with a relevant
  metaphor
• Myers People prefer progress indicators help to
  predict completion time of the task
• Aaron Marcus Principle of color economy use of
  maximum of 5/-2 colors
• Thomas S. Tullis Use of color to group
  information

7
Treatments
8-color
3-color
Island
Bar
Questionnaire Screen
8
The Experiment

• Independent Variables
• Layout
• Amount of Color
• Dependent Variables
• Time to correct completion
• Error rate
• Retention time and accuracy
• Subjective Preference
• Design
• 2x2 between groups
• Subjects
• 32 Junior Senior Computer Science Majors at UMD

9
The Hypotheses

• Null Hypothesis There will not be any
  statistically significant difference in the
  measured efficiency between the four versions.
• Hypothesis I Users will answer questions more
  rapidly and with fewer errors while using the
  progress-bar interface as compared to the
  original island-based interface.
• Hypothesis II Users will answer questions more
  rapidly and with fewer errors while using 3-color
  interface as compared to the 8-color interface.
• Hypothesis III The retention of information will
  be better with progress-bar layout using 3-colors
  as compared to the other layouts.
• Hypothesis IV Users will prefer the 3-color
  progress-bar interface to the other 3 interfaces.

10
Results
11
Results (contd.)

• No statistically significant differences in user
  performance
• Statistically significant difference between
  versions with different color
• In the 3-color versions, users preferred the bar
  layout
• In the 8-color versions, users preferred island
  layout

12
Discussion

• Possible reasons hypothesis was not proven
• Subject base (Junior and Senior CS majors)
• Our new version did not have an optimal
  interface--we changed only color and layout
• Subjects used only one version, so subjective
  satisfaction data was highly varied due to high
  variation of user expectations--subjects had
  nothing to compare the application to

13
Conclusion

• Layout and color had no measurable effect on
  performance.
• When number of colors were reduced, users
  preferred the new layout.
• Future research
• Will a within-subjects design show a preference
  for the 3-color bar layout over all other
  versions?
• Is text more effective than either GUI layout?
• Would a different metaphor be more effective?