Product and innovation · Short course
The Two Gaps
User experience, or UX, is how a screen feels to the person using the screen: how quickly the person works out what to do, and how sure the person is about what happened. This short course teaches UX in six lessons of about five minutes each. Every lesson starts with a small experiment that runs on you, so you feel the effect before you read the research behind the effect. Lessons 1 and 2 are open now, and lessons 3 to 6 are being written.
The model
Every screen asks people to cross two gaps
Don Norman, a cognitive scientist, described two gaps between a person and any system the person uses (Norman, 1986, and The Design of Everyday Things, 1988). The first gap lies between what a person wants and what the person can see to press. The second gap lies between what the screen shows back and whether the person can tell that the task worked. Norman calls the two gaps the gulf of execution and the gulf of evaluation. This course calls them the gap of doing and the gap of knowing.
Good design narrows both gaps. Each lesson below works on one gap, and lesson 6 covers screens that narrow a gap for the company at the customer’s expense.
Lesson 1 · Gap of doing
More options, slower hands
Hick’s law is the finding that a choice takes longer as the options grow, by about the same extra time each time the number of options doubles.
- Try it
- Your result
- Read with care
- At work
- Today
Try it
A word will appear above a row of buttons. Tap the button that matches the word as fast as you can. Round one has two buttons and round two has eight buttons. Each round has seven taps, and the first tap in each round is practice, so the first tap does not count.
Press Start when you are ready.
Your result
William Hick, a British psychologist, timed people pressing one of up to ten keys when one of ten lamps lit (Hick, 1952). Ray Hyman ran a similar test with up to eight lights (Hyman, 1953). Both found that reaction time rose by a roughly fixed step each time the number of equally likely options doubled. Each added option therefore costs less time than the option before, but the cost never reaches zero.
Read with care
Where Hick’s law stops
- Hick and Hyman measured simple choices between options that were equally likely. In our view, the law fits a decision between options that look alike, such as fourteen identical buttons. The law fits less well when a person searches a list for a word the person already knows, such as their own province.
- Fewer options is not always better. A widely cited study found that shoppers offered 24 jams bought less often than shoppers offered six (Iyengar and Lepper, 2000). A review of 50 experiments with 5,036 people found an average effect of more options of “virtually zero”, with large differences between studies (Scheibehenne, Greifeneder and Todd, 2010).
- Removing an option that people need moves the work to a phone call or another screen. Hiding options behind a menu adds a tap for each option. Aim for a clear first choice with every option still within reach.
At work
Agents in your broker portal must pick one of 14 coverage add-ons before the portal will price a home quote. The screen shows all 14 add-ons as identical buttons, and agents spend about 40 seconds on that screen. Which change is most likely to shorten the 40 seconds?
Today
Open one screen your team uses every day and count the options on the screen at the moment of the main decision. Then ask the team which option is chosen most often, and check whether the screen puts that option first.
Lesson 1 is done. Your result is on your course card at the end of the page.
Lesson 2 · Gap of knowing
The same wait can feel shorter
Perceived wait is how long a wait feels, and a screen can change the perceived wait without changing the clock.
- Try it
- Your result
- Read with care
- At work
- Today
Try it
Press Play and watch two progress bars fill, bar A first and then bar B. Then say which bar finished faster.
Which bar finished faster?
Your result
In a study of 22 people, each person watched nine progress bars that ran for 5.5 seconds. Bars that moved fastest near the end were seen as faster than a steady bar. Pauses made a bar feel longer, and a pause near the end weighed most (Harrison, Amento, Kuznetsov and Bell, 2007).
A second line of research looks at what the screen says during the wait. In five experiments on simulated travel and dating websites, people sometimes preferred a website that showed the work being done on their behalf over a website that returned the same results at once. The researchers call the effect the labour illusion (Buell and Norton, 2011).
Read with care
A shorter wait still beats a better-looking wait
- Jakob Nielsen gives three limits for how fast a screen should respond: about 0.1 seconds feels instant, about 1 second keeps a person’s train of thought, and about 10 seconds is the limit for holding a person’s attention (Nielsen, Usability Engineering, 1993, drawing on Miller, 1968, and Card, Robertson and Mackinlay, 1991). Shorten the wait first, then shape what the person sees during the wait that remains.
- Show only real work. Buell and Norton studied screens that signalled real effort. In our view, a screen that invents steps misleads the customer, and the trust lost when the trick is noticed costs more than the seconds saved.
- A popular story says Houston’s airport cut complaints about baggage claim by making passengers walk six times further, so the passengers spent less of the eight-minute wait standing still (New York Times, 19 August 2012). The story comes from a newspaper column, so treat the story as an illustration.
- Skeleton screens, grey outlines of a page that appear while the content loads, are common practice. We have not found a peer-reviewed study comparing skeleton screens with spinners, so treat any claim that one beats the other as opinion.
At work
A home quote in your agents’ portal takes nine seconds to price, and the portal shows a spinning circle during the wait. The pricing team can bring the wait down to six seconds this year, and no further. Agents tell you the portal is slow. Which change is most likely to reduce the complaints?
Today
Time the slowest screen your customers use with the stopwatch on your phone. If the wait is longer than one second, write down what the screen shows during the wait, and whether a customer can tell how much of the wait is left.
Lesson 2 is done. Your result is on your course card at the end of the page.
Apply it
Take the two gaps to your next demo
The lessons teach why screens work on people. To check a real screen, use Looks Right, Feels Wrong, which turns twenty design principles into five questions you can ask of any screen your team or vendor builds.
Your course card
Two questions for the next demo
- Doing: at the main decision on this screen, how many options does a first-time user face, and which option does the screen put first?
- Knowing: during the slowest wait, what does the screen show, and how does the person know the task worked?
A screen works when a person can see what to press and can tell what happened. Every lesson in this course narrows one of those two gaps.
Seeing how technology lands on people is one part of digital intelligence, and the 30-day course builds the rest of that skill one day at a time.
Sources
- Donald A. Norman, “Cognitive engineering”, in Norman and Draper (eds.), User Centered System Design, 1986, and The Design of Everyday Things, 1988 (revised 2013), for the gulfs of execution and evaluation. Norman credits the gulf idea to Jim Hollan and Ed Hutchins (Interaction Design Foundation summary).
- W. E. Hick, “On the rate of gain of information”, Quarterly Journal of Experimental Psychology 4(1), 1952, pages 11 to 26.
- Ray Hyman, “Stimulus information as a determinant of reaction time”, Journal of Experimental Psychology 45(3), 1953, pages 188 to 196.
- Sheena Iyengar and Mark Lepper, “When choice is demotivating”, Journal of Personality and Social Psychology 79(6), 2000.
- Benjamin Scheibehenne, Rainer Greifeneder and Peter Todd, “Can there ever be too many options? A meta-analytic review of choice overload”, Journal of Consumer Research 37(3), 2010, pages 409 to 425.
- John Carroll and Mary Beth Rosson, “Paradox of the active user”, in Interfacing Thought, MIT Press, 1987.
- Chris Harrison, Brian Amento, Stacey Kuznetsov and Robert Bell, “Rethinking the progress bar”, ACM UIST 2007 (paper).
- Ryan Buell and Michael Norton, “The labor illusion: how operational transparency increases perceived value”, Management Science 57(9), 2011, pages 1564 to 1579.
- Jakob Nielsen, “Response times: the 3 important limits”, from Usability Engineering, 1993 (Nielsen Norman Group).
- “Why waiting is torture”, New York Times, 19 August 2012, for the Houston airport story.
- The names “gap of doing” and “gap of knowing”, the two experiments, the broker portal scenes, the answer feedback and the course card are our own.