This is the first article in a series on common usability and graphical user interface related terms. On the internet, and especially in forum discussions like we all have here on OSAlert, it is almost certain that in any given discussion, someone will most likely bring up usability and GUI related terms – things like spatial memory, widgets, consistency, Fitts’ Law, and more. The aim of this series is to explain these terms, learn something about their origins, and finally rate their importance in the field of usability and (graphical) user interface design. We start off with spatial memory – my personal favourite.
The two scientific articles linked to in this story might not be accessible for you. You need a subscription to the distributors of these articles in order to read the full-text versions. Students can usually access these articles either by using a computer at the university campus, or by setting up a proxy at home. Contact your university’s IT dpt. for more questions.
What is spatial memory? Memory is a given, but what does spatial refer to? According to the Concise English Dictionary, the definition is as follows:
“Spatial or spacial adj relating to or occurring in space.”
Spatial memory, therefore, relates to how the brain stores information regarding the location of physical objects in space – the environment around you. Spatial memory is extremely vital for many species’ survival, including us humans: it allows you to find your way in a familiar city, it allows a rat to learn the location of food in a maze, and it allows cats to find their food bowl in the kitchen. Obviously, earlier on in our evolution, it allowed our ancestors to locate sources of water, food, and shelter.
On a side note, research (Kessels, 2002) has shown that spatial memory is not specifically tied to the right hemisphere of the brain (the left hemisphere for some left-handed people), as was thought (and proclaimed in popular psychology). The theory now is that spatial memory consists of a set of sub processes (and of course, as is usually the case in psychology, spatial memory itself is a sub-process), some of which take place in the left hemisphere, and some in the right hemisphere. Just in case you want to know.
A good, real-world example of spatial memory is the tea box I have in my kitchen. It is a wooden box with a lid, and the inside of the box is divided into 12 squares, each of which carries a specific tea – Earl Grey, traditional English tea, Ceylon, mango, cinnamon, you name it. My tea box is located in one of the top kitchen cabinets, and seeing I am not a very tall guy, I cannot look into the box without taking it out of the cabinet. Now, over time, spatial memory has allowed me to learn in which square each of the individual types of tea are. Without specifically sitting down, with the box in front of me, learning each tea’s location, spatial memory has gradually taught me where each type is. I do not have to take the box out if I want cinnamon tea – I know, by spatial memory, which square I need to put my fingers into.
Now, how does spatial memory apply to the world of graphical user interfaces? Take the application menubar for example. Just look at a set of screenshots of various menubars in Windows XP (Media Center Edition 2005, in case you want to know):
As you can see, and, as you probably already know, menubars in one operating system (and, in fact, even across various operating systems) are fairly consistent, at least for the first three to four entries. Whenever an application breaks the usual menu arrangements, like one of the examples above (that is Notepad++, in case you want to know), it stands out, simply because it does not follow the usual pattern. If we go one level deeper, into the individual menus, you will notice that even those tend to be arranged in a consistent manner (where possible, of course).
Spatial memory is one of the prime reasons why a consistent user interface goes a long way in making an interface easy to use. If a user is confronted with a new graphical user interface for the first time, it really helps if applications follow a consistent menubar layout, menu layout, and so on. The user will not be forced to learn specific layouts for each individual application; after having used applications A, B, and C, the brain will get around in menus in application D much faster if they follow the same layout as A, B, and C. That may seem like elementary knowledge, but I still see a lot of people calling consistency freaks like me whiners. When a user complains of an inconsistent user interface – they are definitely not whining.
There is one fairly obvious side note to the importance of spatial memory in relation to graphical user interfaces: isn’t it the case that spatial memory in humans is tailored for a 3D environment? This becomes even more the case as soon as you realise that our ancestors were moving up and down trees, putting additional importance on the third dimension. In other words, how useful is the human spatial memory in today’s 2D interfaces?
An interesting study performed in New Zealand gives some very, very interesting insights into this particular question (Cockburn & McKenzie, 2002). Cockburn & McKenzie set up various virtual, as well as real world interfaces, which varied the ability of test subjects to use the 3rd dimension. “The aim of the experiment is to investigate differences in people’s ability to use their spatial memory in physical and virtual systems, and to see what effects occur as richer levels of a third dimension are available.”
The results were, to me, unexpected. “Results show that the subjects’
performance deteriorated in both the physical and virtual systems as their freedom to locate items in the third dimension increased. Subjective measures reinforce the performance measures, indicating that users found interfaces with higher dimensions more ‘cluttered’ and less efficient.” In other words, the test subjects’ difficulty with spatial memory related tasks increased as the interfaces went from 2D, to “2.5D“, to 3D – not only did their reaction times increase, their assessment of the task was more negative as well.
In other words, users do, in fact, benefit from spatial memory for 2D tasks, despite the fact that you would think that the human spatial memory system would be tailored to 3D tasks. This is probably also one of the prime reasons why there seems to be little incentive in the computing world by companies like Microsoft and Apple to move to true 3D interfaces – I am sure their usability testing came up with similar results as the New Zealand study.
Conclusion
How important would I rate spatial memory when it comes to usability and graphical user interfaces? Considering that it is a building block (or a sub-process if you will) of not only learning in general, but also specifically of learning the locations of items (both physical as well as virtual), I would rate it as a very important aspect, something user interface designers should really take into account. Seeing graphical user interfaces are all about allowing the user to access [find] functionality [objects] in an interface [space], spatial memory is a vital cognitive element.
If you would like to see your thoughts or experiences with technology published, please consider writing an article for OSAlert.
Yes, spatial memory is VERY important when it comes to GUI. I drive myself nuts over something as seemingly simple as 3rd party utils having their own entries in context menus in Windows – TuneUp for example places its Undelete command right above the separator in Trash bin’s context menu and I constantly find myself clicking it instead of “Empty trash”. And that’s just a little insignificant example. Thanks for a nice article, Thom!
one more person moaning about Fitt’s Law I think I’m going to spoon my eyeballs out.
Yes, most of us who are in development know it.
Yes, we all know the Windows start button sucks at it.
My problem with it is that it’s over-applied to situations in which it has no application by people who have little to no context to what they’re speaking about. It drives me absolutely nuts when people start spouting about GUI design without ever having completed any guided studies on UI design, been involved in TA testing of GUI principles or prototype software, and generally think UI design is about ‘teh webpage should never scroll’.
Gah.
So thanks for trying to put something together that is more in-depth and might actually enlighten people that there’s more to UI design than freakin’ Fitts.
Edited 2007-10-25 19:09
Definitely agreed. I find it especially silly when Fitt’s Law gets cited as some sort of trump card argument “proving” the superiority of E.g. global menu bars. Fitt’s Law, IIRC, just has to do with where UI elements can be placed to make them easiest to “hit” with the mouse pointer – it doesn’t say a thing about which particular elements should be placed in those easy-to-hit parts of the screen. That’s the difference between usability theory and its practical implementation.
If you are not using a mouse but a touch screen instead does Fitt’s Law still apply? I wouldn’t think so because the mouse pointer stops moving when it gets to the edge of the screen whereas your hand does not. I would think that with both mouse and touchscreen interfaces that the size of the widgets is more important than Fitt’s Law anyway. Consistent spacial layout would come before Fitt’s as well. Who cares if it is hard to reach for a button if the button is not in the same place every time anyway.
i dont know if there was any experience data collected on it but i recall that in the 94 winter olympics the computer systems where supposed to be touch screen controlled.
and i kinda agree with you, the problem with the mouse is that we do not have awareness of its locations at all times. hell, i know i have the habit of spinning it in circles when it need to find it before i do anything.
with our hands we have built in awareness, thats why we can in theory reach out of view and still have a good idea of where our hands are.
i also wonder what a general feedback system on a mouse would be like. say having the left button jump each time one move over the edge of a ui element.
right now, our sense of where the mouse is is based to much on sight. we cant use hearing, smell, touch or anything like that to tell where the mouse is in relation to other things.
Fitts’ Law doesn’t say “put a menu at the top of the screen”, but it does explain why placement a is superior to placement b, as long as you buy into the whole “infinite height” thing.
True, but it could be applied to any user interface element just as easily as menu bars.
The start button doesnt suck at it, it is in one of the four easiest to hit places on the desktop, a corner. You can have the mouse anywhere on the screen, close your eyes, and still be able to trigger the start menu.
Uhm… Doesn’t work here… The start button is a couple of pixels from the left and a couple of pixels from the bottom of the screen. This is windows xp with the classical theme.
Although I can trigger the start menu with my eyes closed (using the windows button on the keyboard).
Interesting. I’m using the classic theme too, and here it works just fine.
+1 (Me too)
Prior to windows XP, it didn’t work.
In XP, if you ram the mouse to the corner and click, the mouse pointer will actually be -repositioned- without you moving it!
if anything needs a clear and rugged definition it is usability.
PS; it is not the (magical) sum of its parts.
“[Usability refers to] the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use.” – ISO 9241-11
I got that by going to google and define usability.
The third link down was
http://www.upassoc.org/usability_resources/about_usability/definiti…
Who would have thought that it was an International Standard.
Googleing “ISO 9241-11” brings up more information.
Nice to have that in a standard. One thing that I have noticed is that most articles on usability actually focus on “learnability” (how easy it is to use an interface you have never seen before). I hope that this series will not be yet another one of those.
I like to show the difference between learnability and usability by comparing Notepad and Vim. None has ever said that Vim is easy to learn, Notepad on the other hand is very easy to learn. But when it comes to actually using those programs to perform some text editing (such as modifying source code) an experienced user will perform the task much faster and more efficient in Vim than in Notepad.
A consistent UI is great for learnability but it is not necessarily good for usability. That is simply because some common UI principles are not well matched to certain tasks and will force the user into inefficient procedures (which can be easy to learn but which takes more effort to perform).
I wrote a litte text a while ago explaining that sentence. Might be interesting:
https://wiki.ubuntu.com/Usability
Edited 2007-10-27 23:10
This definition of usability is based on arbitrary concepts such as ‘goals’ and ‘satisfaction’. This is never going to develop into repeatable or transferable methodologies.
What makes a good sofa, alarm clock or UI ? Who can tell.
You should leave interface design to be the craft that it is; with its market, traditions and apprenticeships. Perhaps one day it will get the respect it craves (and deserves) instead of the derision it so openly invites.
The (in?)famous ArsTechnica article on the “Spatial Finder”:
http://arstechnica.com/articles/paedia/finder.ars/2
Thanks, IBM.
http://en.wikipedia.org/wiki/Common_User_Access
did said test about 3d space navigation involve true 3d, as in virtual reality, moving around in the world and so on?
or did it involve someone sitting with a mouse and keyboard and try to interact with a 3d world on the other side of a 2d display?
while i cant prove it, i have a feel that there could be a difference between the two.
still, one could maybe say that our world is at best a 2.5d world. as in, we cant really float into the air at will, so for us its two dimentions (forward/back, left/right) with at times some kind of height data added as a secondary condition (upper or lower shelf and so on). so when we move from tree A to tree B to get some fruit, we first move in 2d space down tree A, then in 2d space between the trees, and in the end moves in 2d space up tree B.
a bird on the other hand maybe be more adept at finding things in 3d…
btw, i would love being able to switch of my spatial memory when dealing with menus. more often then not i click and then find out that my memory was faulty or maybe the coordination between it and the movements done where not perfect…
Edited 2007-10-25 23:18
That’s why they used not only 3 virtual environments, but also 3 real world ones. In both cases, the results were clear: adding a third dimension makes it harder for people to locate items.
ah, i guess thats what i get for being unable to access the article itself…
consistancy is important sure, but if every application is required have the same user interface then there will not be advancements in user interfaces.
The web is a great example of how users amazingly manage to navigate user interfaces that are completely different from each other without much trouble.
I think it is important to recognize that a user may make a distinction between a web interface and the user interface of a program installed on the user’s computer.
Web sites are “out there” whereas local apps are “here”.
A user may reasonably expect that everything that is “here” is known and familiar while not having that expectation for anything that is “out there”.
“The web is a great example of how users amazingly manage to navigate user interfaces that are completely different from each other without much trouble.”
Er… without much trouble ? Well, maybe we’ve seen different sets of people then.
:-P, Thom…
Nice article to read, but too much emphasis on one thing will make it a bit boring.
Anyway, on the thing about deteriorating spatial memory with the increase in dimensions is more likely tied to 2 things. Logically, it is much more difficult to design good 3d interfaces than 2d ones, which will undoubtedly hinder the use of spatial memory. Referring to keyboard and mouse especially, it is difficult to traverse the 3rd dimension. Regular monitors are also just plain 2d — no point emulating the 3rd. Unless the time comes when we can have perfect 3d input and output devices, where perfect input refers to high accuracy, precision and huge movement area, and perfect output refers to hologram-like projections, should we talk about it.
Second major factor in spatial memory’s failure to span the 3rd dimension is that out ancestors were just plain lazy. Look at the movie industry. How many action movies portray the hero running around in enemy territory right above the minions’ heads? Head movement is mostly left and right for a simple reason — its tiring to move up and down.
A complete moron would say “people remember where stuff is, so it’s good to put stuff in the same place so people don’t need to go looking”.
It takes the average expert years to develop hard to understand terminology (and many paragraphs of waffle to explain the terminology) just to say the same thing.
I agree. This is a common theme on the web when you got people who have no real experience in anything but want to sound smart when they talk about it. There’s nothing like writing something with a lot of words and air to help these experts reassure themselves that they’re some sort of professional/intellectual. Overthinking and overcomplicating… ah, how it feels to smart.
Edited 2007-10-28 21:48