At BusinessLogs, we help you communicate better with your customers by designing elegant, friendly application interfaces & blogs. Learn More ›

Attentional Spotlighting in User Interfaces

In visuospatial terms, a saccade is an eye movement from one area in the visual field to another, and usability testing using eye-tracking equipment tracks the timing and progression of saccades across a user interface in response to certain stimuli. These stimuli may include 1) the first presentation of the interface to the user, 2) the modification of the visual field due to animation (Flash, banners, animated GIFs, etc.), 3) the slow download of information to be presented, or any number of other things that can attract your gaze.

Unfortunately, where you are looking is not necessarily where you are paying attention. The spotlight theory of attention states that you can attend to a different section of your visual field than you are currently looking at. This means that even though you are currently looking at this line of text, your attentional spotlight could be on your buddy list to see who just signed on, or down at your Mac OS X dock to notice an application icon bouncing up and down, or even on nothing in particular because you are paying attention to a thought process going on in your head as opposed to what’s happening on the screen. Ever read a passage of a book and then realize after the paragraph is complete that you don’t remember anything you just read? That happened because your visual focus wasn’t in sync with your spotlight of attention, thus your attention wandered and you missed the point of your reading.

User Interfaces and Spotlighting

Users who have no familiarity with the interface presented in front of them will pay more attention to what they are doing than if they are a seasoned user with hours of experience. I no longer look at the keyboard to open a new tab in Safari, I simply position my left thumb and index finger unconsciously and a new tab appears. Did I pay attention to the physical act of moving my fingers in a manner that puts them on top of the correct keys? Did I have to break concentration in order to remember the correct keystroke sequence to make a new tab? Of course not, because it is a learned response which requires no attention in order to perform it correctly.

Difficulties arise when performing usability tests on subjects who are very familiar with their scenario and can breeze right through it based on learned interaction sequences. When you stop attending to the interface you are using and simply go on subconscious learned activities, the usability testing (and eye-tracking) is now biased and altered. The interface’s spatial cues are there to inform the user about possible interaction paths, but when your attentional spotlight is not focused on the task at hand, the interface’s spatial cues are bypassed and you are not interacting with the UI in the manner it was designed. Your familiarity with the interface is effectively blocking positive interaction, thus usability testing needs to be adjusted.

Combatting Learned Interaction

When testing seasoned users is a necessity, think aloud protocol should be used in order to facilitate a full interaction with the interface. By forcing one to speak about their cognitive processes during testing it verifies that they are not using learned interactions. When both talking and doing at the same time, you are increasing the chances that their spotlight of attention is fixed at their gaze point instead of them attending elsewhere. Now that they are focused on the task at hand, they will pay more attention to the user interface in front of them and can turn back into a valuable usability tester.

And Now What?

People sometimes forget that the user experience industry has its base in psychology, and that knowing how information is processed in the brain helps someone become a better user interface designer.

About Mike Rundle

Comments

  1. Aby Rao says:

    This is a very interesting piece of information, that corroborates several practical issues in UI.

  2. SteveR says:

    “Of course not, because it is a learned response which requires no attention in order to perform it correctly.”

    I guess I don’t understand what you mean by “attention”. My mind would have to decide that I needed a new tab opened, then it would have to inform my arm/hand/fingers how to achieve that. My body does do things without me thinking about it consciously, such as breath and digest, but I would argue that opening a new tab does require my attention.

  3. SteveR: I believe your missing the point, or your just playin’ with our heads.

    The fact is that opening a new tab, for Mike, is now a conditioned response to the need for one. Therefore it does not require any conscience effort to open one.

    I’m sure he used the word “attention” loosely in this regard.

  4. Mike Rundle says:

    Hey Steve-

    What I meant was that your decision to open a new tab doesn’t divert your attention that is focused on the interface at all. The split-second cognitive process that takes place doesn’t move your attentional spotlight from where you are currently attending, thus the amount of attention diverted from the task in order to make the decision is proportional to the amount of familiarity you have with the task.

    If you’ve never opened up a tab before, your attention will be focused on the keyboard in order to hit the proper sequence of keystrokes, however the more you learn that action the less you need to attend to it while you perform it, freeing your attention to stay focused on the interface instead of the keyboard.

  5. Chad Lapa says:

    Steve – to go along with what Mike said, try and think of it on a simpler level then opening up a new tab. For example, take the amount of attention diverted to typing up this message? My attention is not focused on where to move my fingers in order to correctly hit the key I want in order to correctly spell out my message. I am actually thinking about what I want to say and my fingers are moving automatically. This cognitive process of typing is like Mike stated not enough to divert my attention away the actual message that I am trying to convey. If I had to divert my attention to every letter that I typed and what fingers to move where, then it would take a much longer to time to convey my point then the 1 minute it actually took me to type it :-)

  6. Steve S says:

    How do you reconsile “think aloud protocol” with the idea in Blink (Gladwell) that there are things we know but can not express, and when we do try to express them, we actually reconcile our thoughts to what we were able to express? Wouldn’t you need to avoid these experienced users all together?

  7. Ron says:

    Maybe there comes a time when “information architecture” needs to relax, and stop over engineering a user experience for everyone. Some things you just can’t account for. Like what your visitors are thinking.

    Other than that, this is interesting stuff.

  8. Mike Rundle says:

    Steve – I definitely agree with you. My post was simply describing why testing seasoned users is probably a bad idea. Does it still happen? Absolutely.

    I think the best case for testing seasoned users is when fine-tuning quantitative metrics associated with learner-centered design. Students using software like Blackboard are used to the typical interfaces and templates associated with the app, however when teachers and professors put in their own information, they now have to adapt to new information provided in the same interface.

    Same car, different interior.

    The usefulness of testing seasoned users in this manner is that you are now testing the information presented in the application as opposed to the interface of the app itself. Testing new product releases is similar to this — like when usability testing the new Keynote over the old version because Apple doesn’t want users to don’t know anything about the first version to test the second one, they want people who are familiar with the old interface and can comment on the usefulness of the changes made to the second one. If a dev team wants to find out how they’re improved over a previous version, testing brand new users may not be the avenue to drive down.

    Just some of my thoughts. Different user-centered design methodologies and usability testing techniques can apply in various situations, so it wholly depends on the context of the testing.

  9. Like with everything in this field, it’s best to start with a hypothesis; a goal to find out something in particular. Then you can look to what type of testing will get you close to an answer.

    The biggest thing is understanding the different types of testing, the levels of interaction that tend not to influence you participants too much, etc.

  10. Peter Brawn says:

    It’s true that you could be looking at something but not ‘paying attention’ to where you are looking – like daydreaming! In that sense visual attention requires some effort and yes, you’d expect people who are more familiar with an interface to behave differently to people who are less familiar (e.g. differences in search ‘efficiency’ that would be reflected in their scan-paths).

    However, it’s actually really hard to deploy visual attention to something that you’re not looking at (visual attention is a difficult thing to be introspective about – but try it). Can you really say much about the items in your task bar or buddy list – while fixating something in the middle of your screen? What about away other items in your visual field away from the screen? Why would you? Acuity drops off rapidly for items that appear in peripheral vision and it’s worth keeping in mind that one of the main purposes of visual attention is to guide eye movements.

    There are more recent models of visual attention available that go beyond this basic spotlight metaphor – for example, see the work of Jeremy Wolfe and John Duncan. For the interesting question as to whether we actually ‘see’ anything at all in the absence of focussed visual attention – there are some great demonstrations out there by Daniel Simons and Ron Rensink.

  11. Paddy says:

    I agree with Peter Brawn, and would add that we can attend to items bouncing in our OS X dock or buddies signing on without moving our eyes is that they are designed precisely to achieve this – they use movement to exploit the strengths of our peripheral vision. This wouldn’t apply in many other cases.

    Also, “Blink” was mentioned – a thorougly enjoyable book. A great, older paper on this sort of thing is “Telling More than We Can Know: Verbal Reports on Mental Processes”, by Richard Nisbett and Timothy Wilson, Psych. Review 84/3 (1977). There was also a rebuttal to this by Ericsson and Simon, same journal, later issue.

Trackbacks

  1. With the User Experience, Small Things Add Up

    I’ve been an Apple geek for all of my life, since I first wrote: 10 home 20 Matthew is so cool!!!!!!!!!!!!!!!! 30 goto 10 on an Apple II, thus amazing my friends with how cool!!!!!! I was. When I made

Speak Your Mind

*