|InfoVis.net>Magazine>message nş 124||Published 2003-06-24|
|También disponible en Espańol|
The digital magazine of InfoVis.net
We, computer users, are very used to the notion of menu, which is strongly associated to menu bars, pop-up or drop-down menus and dialogues of the user interfaces we deal with every day.
Most of them are examples of linear menus, where the options are laid out in a linear fashion, vertically or horizontally, i.e. the options are one under the other or side by side.Â
Unlike in linear menus, pie menus lay out the options circularly as if they were pie slices, which gives them their name.Â
This is an infrequent type of menu although by no means a new one. Already in 1988 Callahan* and other researchers were comparing pie menus against linear menus on 33 people. They found the that former ones were an average of 15% faster regarding their ability to be located by the user. Nevertheless linear menus were only worse in 50% of the options that lied farthest from the initial point of cursor.
Why, then, havenâ€™t we seen them except in some applications like video games or prototypes of advanced programs like DENIM? (see issue number 122)Â
Two weeks ago, during a course held in Stockholm for young professionals of digital design, I asked the students if they thought that a linear menu could be more or less effective than a pie menu. The answer was unanimous and almost immediate. All of them were prone to think that the best by far was the linear one. My own reaction when I wrote the article on DENIM was the same. So used we are to the linear menu, to the â€śofficial paradigmâ€ť.
Jason Hong has a page on this type of menu, from which you can download Java programs, including its corresponding source code. According to him there are three main advantages:
One of the problems of pie menus is that, when clicking on an option, if it has sub-options, they open up, partially occluding the former one, which doesnâ€™t occur with linear menus. A solution to this problem is semi-transparent menus that eliminate the occlusion.
So we come back to the question: if they are more efficient, like the â€śFasteroids game lets you test, why donâ€™t they dominate instead of linear menus?
Iâ€™ve asked this question to Don Hopkins. His answer deserves a complete issue of this newsletter. Next week weâ€™ll have it.
Jack Callahan, Don Hopkins, Mark Weiser and Ben Shneiderman. â€śAn empirical comparison of pie menus versus linear menusâ€ť Proc CHIâ€™88 Human Factors in Computer Systems. Chapter of ACM (March 1982), 190-196
Links of this issue:
Subscribe to the free newsletter