SHORE 2001 Logo
SHORE 2001 Logo University of Maryland Logo
Student HCI Online Research Experiments
Abstract
Introduction
Experiment
Results
Discussion
Conclusions
Acknowledgements
References
Appendices
Credits
Feedback
SHORE 2001 : Web : The Menu Design and Navigational Efficiency of the E-Maryland Portal

Results


 

Chart 3.1: Graph showing average time to correct completion for each task, grouped by which version of the site.  In general Static-Shallow had the fastest average times, except for task 5.  Below is the chart of data.

 

 

Dynamic-Deep

Static-Deep

Static-Shallow

 

Average (sec)

Std. Dev.

Average (sec)

Std. Dev.

Average (sec)

Std. Dev.

Task 1

20.25

13.6

29.00

7.2

11.50

9.4

Task 2

50.00

28.4

44.00

51.1

19.13

7.3

Task 3

46.50

28.9

49.50

58.8

11.25

11.5

Task 4

29.63

36.3

47.25

35.8

17.63

11.9

Task 5

18.63

12.8

6.13

1.4

19.50

23.9

Task 6

127.75

63.7

63.75

60.3

15.25

8.4

Task 7

140.13

39.4

152.13

56.0

83.75

46.5

Task 8

19.38

11.5

12.25

7.6

28.75

15.0

Task 9

58.25

32.3

103.88

80.0

53.75

37.6

Task 10

89.75

70.8

59.25

56.9

35.50

45.4

Task 11

44.00

48.7

59.00

36.1

21.75

23.1

Task 12

95.50

74.4

101.38

63.7

31.25

23.6

Table 3.1: Shows the average times and standard deviation it took subjects to find a link under which version.

 

 

Dynamic-Deep

Static-Shallow

Task 1

N

Y

Task 2

N

N

Task 3

N

N

   Task 4

N

Y

Task 5

Y

N

Task 6

N

Y

Task 7

N

Y

Task 8

N

Y

Task 9

N

N

Task 10

N

N

Task 11

N

Y

Task 12

N

Y

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3.2: Indicates if the results are statistically significantly different (p < 0.05) for tasks compared with Static-Deep.
These results show that Dynamic-Deep was never faster than Static-Shallow and was slower for only task 5.
Static-Shallow was faster than Static-Deep for tasks 1, 4, 6, 7, 11, and 12, but slower for Task 8.

 

We used the one way ANOVA to test each task if the results were statistically significant based on the critical values.  Dynamic-Deep and  Static-Deep were compared for the menu interface.  Static-Deep and Static-Shallow were compared for link hierarchy.

 

Since there is only one task (5) that Dynamic-Deep and Static-Deep had that indicated a statistical significance, we assume that Static-Deep has no performance advantage over Dynamic-Deep in finding a link.  On the other hand, Static-Shallow and Static-Deep had results for a little over half the tasks that indicate some statistical significance, so we assume that Static-Shallow has some performance advantage over Static-Deep.  And since Static-Deep has comparable times with Dynamic-Deep, then we also assume that Static-Shallow has some performance advantage over Dynamic-Deep.

 

 

Recall Task Data:

 

 

Number of People

Lifelines

8

Business Services

5

Maryland Scenes

2

Citizen Services

2

 

 

 

 

 

 

 

 

 

 

 

Table 3.3 : Shows the number of people who could recall correctly the exact name of the main category.

 

Only one subject had correctly recalled all four of the main categories.  Many subjects, however, could recall the idea of the category or some permutation of the name.  For example, 19 subjects recalled business(es) or business info as a category and for citizen services 13 subjects put community or a phrase with community in it.  Some other categories that subjects came up with include: lifestyles, life guide, recreation, life in MD, Maryland citizens, around and about town, services for citizens, travel, transportation, education, Maryland resources, family, individual, Maryland places to see, citizen/community and life.  In comparing the total number of correct categories recalled, Static-Shallow had 9 total, Dynamic-Deep had 5 and Static-Deep had 3.  The recall error rate is the number of incorrect categories the subjects named when asked to name the 4 main categories.

 

 

Dynamic-Deep

Static-Deep

Static-Shallow

  Recall Error Rate

87.5%

90%

82%

 

 

 

 

Table 3.4: Shows the recall error rate when asking the subject to name the 4 main categories after completing the first task set.  Subjects that used Static-Shallow to complete tasks 1 to 3 seem to have a little better recall rate than Dynamic-Deep and Static-Deep.

           

 

Subjective Data:

 

wpe4.jpg (7166 bytes)wpe2.jpg (10966 bytes)

wpeF.jpg (9464 bytes)

 

 

Chart 3.2: User Preference - shows that most users prefer Static-Deep, then Static-Shallow.

 

Chart 3.3: Ease of use - shows that most users found the static, shallow version easier to use.

 

Chart 3.4: Most Visually Appealing – even though 50% of the users find that Dynamic-Deep is the most visually appealing it ranks behind Static-Shallow on user preference and Static-Shallow ranks last in being considered visually appealing.