Discussion

The results showed that there was no performance advantage in using the static menu (Static-Deep) over the original E-Maryland menu (Dynamic-Deep) since most of the average times to correct completion were around the same.  It did confirm though that link hierarchy has a large influence on performance times since six of the twelve tasks had statistically significantly faster results with Static-Shallow then with Static-Deep. (We know a result is statistically significant if the F value is greater than the Fcrit value (reference Mike Stark’s page).  

Even though Dynamic-Deep and Static-Deep had comparable times, they did not have the same preference rate.  The reason for this can be found in some of the comments from subjects:

• “In Dynamic-Deep, not being able to see all the menu options at the same time was bad.”
• “too frustrating to navigate thru.”
  “Dynamic-Deep was kinda of hard to look at because you can’t really move the mouse too much or else the menu would disappear”
• “Mouseover of icons lead to results that confused me, should mouseover and have menu appear in same quadrant”

 
The results showed that Static-Shallow had quicker average finish time for the tasks. Besides the argument that broader shallower trees are better, another reason that Static-Shallow has better average times was that many of the subjects knew about the Find function and when given the exact link name, it was easy to locate. One thing interesting to note is that in Figure 3.4 in the results page, 50% of the users found the Static-Shallow easier to use. This may also have to do with the Find function in the web browser.

Static-Shallow did not do as well in preference or visual appeal as Static-Deep  mainly because the huge number of links.  One subject said “Whoa” when they first encountered the secondary page on Static-Shallow.  Another commented that it has “too many options. Can’t view all at once”.  Huge number of links leads to longer pages and longer pages lead to scrolling, and scrolling is considered bad.

The results also indicated that subjects that used Static-Shallow first had a slightly better recall rate of the 4 main categories.  One possible reason for this might be that the users have to remember less steps (clicks) , less paths from one link to another and therefore has more room to remember the main categories.