“This got really creepy really fast. I don’t feel safe.”
Creating uncomfortable user experiences through distortion of the familiar
(Originally published 3/28/14)
The quote in the title is from a reaction to Byzantium Tests.
Last year I stumbled across an article I haven’t been able to forget: Uncomfortable User Experience. It’s essentially about crafting experiences that are bad on purpose, experiences that are unintuitive, confusing, disorienting, and disturbing.
Who would do this to a user? And why? The authors state that they’re advocating for the creation of such experiences in order to shift “the focus of user-experience design from the traditional usability goals of learnability, performance, and minimizing errors to new ones, like fostering emotional and aesthetic engagement.”
Certainly there is a time and place for this uncomfortable user experiences. When I’m checking my email I would rather not be confronted on my adherence to societal norms and expectations, I would simply like to check my messages. And if the Pizza Hut app began questioning my allegiance to our corporatist culture, I would be slightly delighted but mostly annoyed because this political engagement would interfere with my prime objective of ordering pizza. The article addresses where we currently see negative experiences used in a purposeful way, and we can use this begin to think about where else we could introduce uncomfortable UX.
Where uncomfortable UX is used can be broken down by three different goals —
Entertainment. Amusement park rides, video games, and horror films all give users stressful experiences for their own enjoyment.
Enlightenment. We see this in art that pushes your boundaries, and although it may not be pleasant, your eyes are opened to a different view of the world. Also in this category are spiritual practices that tax the body (fasting, abstaining, etc.) with the goal of religious enlightenment.
Sociality. In adverse situations people develop strong social bonds, from the anticipation of waiting in line for a roller coaster together, to the more real danger of bonding through hazing.
So how do we make user experiences uncomfortable? The article describes four ways, key ingredients for the recipes of making people feel bad, and provides many examples from performance and installation art —
Cultural. Affect how the user feels be making a narrative or interface that deals with cultural associations (the work Ulrike and Eamon Compliant puts the user in the role of a terrorist).
Control. The user feels best when they are in control, so subvert this by making them surrender control to the machine (visible circuitry emphasizes the power technology has in the situation).
Intimacy. Make the user uncomfortable by isolating them, or by requiring them to establish intimacy with strangers.
It’s vital to keep in mind that the temporary discomfort of such negative situations must be justified by the longer term goals (entertainment, enlightenment, or sociality). Also we must consider the aspect of informed consent. Is the user aware of what they’re getting themselves into? Are they aware of the risks? If there’s a “point of no return” in the experience, is it explicitly declared? A fascinating side effect of warning users is that the warnings create a sense of danger, heightening the users’ suspense.
Elise Lind wrote a letter to the editor last month responding to “Uncomfortable User Experience” and she critically addresses the fact that all the examples the authors used were physical interfaces and installations. She speculates whether the same tactics can be used in software, hardware, and app development?
And to enchantingly disturbing ends.
We can make people uncomfortable without having control over their physical environment or interface, but it requires creating cognitive discomfort. There are a number of ways we could do this, but what I’m really, really interested in is the mental discomfort that comes from being asked to question your sanity or the validity of your perception of reality: are you really seeing what you think you’re seeing? are you really who you think you are? is there something you don’t know?
We can adopt (and then distort) the conventions of personality tests to create this uncomfortable identity questioning.
Although I have rather… strong… ideas about how personality tests should be discredited and not taken seriously in the workplace or education, when we acknowledge they’re just for fun I think they’re fine. However, even when we are taking the latest ridiculous Buzzfeed quiz, there is still an element of self-examination, causing us to question our psyche and open up to a mindset of self-analysis.
Once the user is in this susceptible mental space, when audio, visual, and emotional disturbances begin to appear in the test, they’re in such a zone where they can embrace the adrenaline and spookiness.
What do I mean by distorted? Let’s look at three examples: one was part of a transmedia advertising campaign; one is a collection of fabulously upsetting tests not associated with an overall narrative; and one is Rorschach-like test woven into a video game.
This video is a preview of just the personality test aspect, which is the only part I’m discussing in this post. There are some walkthroughs of the whole test on YouTube, however the commentators are insufferable.
The test was in five stages. The first was designed by Randy Horton, who we’ll talk about later. That section was typical of his style, beginning with standard personality test fare (“Which image describes your childhood?”) and then deteriorating to more unnerving questions (“How safe are your dreams?”). ARGnet describes the next few stages, “The second phase of testing gives prospective candidates the option to log in through either Facebook or Reddit to respond to emotionally charged questions based on images, while the third phase gives users the option to link up with a webcam to test emotional responses through facial recognition. The fourth phase asks candidates to solve a simple logic puzzle while multi-tasking.”
The final step asked the user to draw a picture that was displayed on their childhood fridge. Subliminal messages along the way influence the user to draw a house, and then the test administrator goes on to show the user how their subconscious was influenced.
It’s inexplicably eerie and powerful, but since we can no longer experience, let’s move on.
Randy Horton, UX designer and art director, helped Campfire with the Byzantium Tests because of his exceptional work on these uncanny gems. Although they’re less polished than the agency work of Campfire, in a way this makes them even more unnerving.
There are three tests (all should come with an epilepsy warning, only the last actually does) —
Go on and play this alone at night.
The introductions sounds innocuous enough, with some neuroscience flavored drivel about shapes, the brain, and “dissociative personality disorder”.
Begin the test and you’re shown shapes and asked to pick which one is lying to you. And which is diseased. And where will you be safe. The experience of taking the test is unsettling.
At the end you’re shown a result which does sort of describe your personality, even though none of the questions were about your personality, thanks to the Forer Effect.
This introduction begins with more neurobabble, this time about schizophrenia. However, this time we get a fantastically upsetting warning—
This diagnostic tool should be administered (and all results intermediated) by an experienced clinician, so that suggestible individuals or those with a precarious sense of self can avoid feelings of depersonalization, loss of affect or ego death.
The test itself forces the user to choose between four squares, each of which contains two oscillating images. The questions start out confrontational, but then dissolve into paranoid nonsense (“Is we stop no it?”). During the test a disconcerting, electronic humming adds to its nerve-inducing nature.
Again, the answers are general enough to fit most people, and focus on negative aspects of personality. However, this time they do not refer to the user in second person, instead referring to them in third person as the “client”. This emotional distance after such a bizarrely intimate test is upsetting.
I think this is the best test out of these three. The disorder it pretends to be about is anxiety, which is far more likely to be relatable to the user than dissociation or schizophrenia. It again refers to the user in an ominous third person, and displays a troubling warning.
The test begins with “reasonable” questions (“Choose the selection which is morally correct”), but then chastises the user for not following instructions.
Eventually instead of questions, we begin to receive statements, “If the darkness touches your skin it will infect you.” A low humming noise and distorted human voices add to the discontent.
The result in in first person again, and includes another warning —
This test could cause a case of information poising or identity sickness. If symptoms of disassociation or existential dread continue for more than a day, please contact a therapist.
These tests are all loosely related, and Horton provides a weak narrative element to them. They include links to some email addresses and websites that could perhaps form a larger alternate reality, but the enjoyable aspect is the disconcerting visceral immediacy of the tests.
Cave! Cave! Deus Videt
It’s as fantastic as it is hard to explain. The trailer doesn’t even try.
It’s a fairly short game, less than an hour.
Go on. I’ll wait.
It’s a beautiful game based on the work of Jeronimus Bosch. Although it deals with all sorts of delicious aspects of art history, medieval medicine, and pop culture, I’m just looking at the psychological test aspects here.
At the very start of the game we hear a clock ticking, and a voice reads the words on the screen —
“The following projective psychological test is provided for educational purpose. Your answers will be recorded and used anonymously for research.”
Something is weird about the voice, it’s been digitally altered. It sounds kind of like the speech from Cooper’s dreams in Twin Peaks — phrases recorded being said in reverse and then played in reverse.
Or, you know, technology.
You can either accept of decline the test. If you accept, you take what looks like a typical Rorschach ink blot test (which are creepy enough on their own). But then you begin to see flashes of distorted images and hear digital static. You can ask the person testing you about what you’re seeing, but they simply dismiss your concerns.
This “subliminal” flashing continues throughout the game, as do the otherworldly voices. It plays into the larger themes in the game of dreams and reality, and starts out the narrative on an adequately sinister tone.
As we can see, sometimes and experience just on the computer can be really unnerving, even without a physical element. Just ask all these people on Twitter that were really, really freaked out by the Byzantium Tests.
The four forms of discomfort addressed in “Uncomfortable User Experience” can be observed in upsetting psychological tests —
Visceral. The flashing, strobe effects hopefully don’t give you a seizure, but may make you physically nauseous or dizzy.
Cultural. Personality tests are a cultural norm of the internet. They are a thing we do.
When they are subverted against our expectations it’s a bit of a shock. We also have cultural expectations to present as mentally healthy, and when it is suggested in any way that we aren’t (seeing things that aren’t there, altered forms of reality, the negative descriptions in the results of the Hypnoid tests, etc.), this makes us uncomfortable.
Control. When the tests asks questions we cannot answer, because they don’t make sense or because they aren’t even questions, the user has a hard time grasping the situation or what to do, and thus loses control.
Intimacy. You use the internet alone. It is intimate.
You will be alone when you take these tests, unlike other psychologically thrilling experiences which you experience with others — theme park rides, haunted houses, etc.
When the Byzantium Tests were live, I wandered onto it unexpectedly. I was on my computer, home alone at night.
It was genuinely frightening.
(But, you know, in a fun way.)
This is what makes the cognitively uncomfortable user experience different — and maybe in some ways better — than the physical, installation-type interfaces discussed in “Uncomfortable User Experience”.
And whether we’re freaking out users for entertainment, enlightenment, or sociality, we should consider the amount of trust users place in their digital devices (and the internet itself) and how this adds an entirely different aspect to vulnerability.
And, oh, how we can use this to create delectable disquieting experiences.