Ambiguity as a Resource for Design – a Review

Reading – Ambiguity as a resource for design, by William Gaver, Jacob Beaver, and Steve Benford

Ambiguity is often seen as a quirk to be avoided, and for good reason. If your results are incomprehensible, then your research may be wasted, but ambiguity can also be used as a pointed tool, getting results much deeper than otherwise expected. When you want your research to be intriguing, thought provoking, and other buzzwords, it may be useful to implement strategies harnessing ambiguity. This paper talks about creating ambiguity in three ways. Ambiguity in information, contexts, and relationships.

Ambiguity in information sounds extremely counterproductive, but hiding some, and not all, of your information can force the interpreter to fill in the gaps in ways that are personal to them. If you’ve ever seen the Mona Lisa by Leonardo DaVinci, you will likely recognize that she does not have a clear expression. This is due to the softening of detail around the mouth and lips. As a result, the viewers are the ones that must interpret the expression themselves, a valuable intellectual task the artwork easily asks.

“Point out things without explaining why…”

Secondly, there is ambiguity of context. Take the idea of a phones ringtone. Straightforward and obvious to its purpose of letting you know someone wants your attention. However, it is just a sound, that can have a multitude of uses. Some mothers download a soothing, soft ringtone, and put the phone near a young child. Calling the phone turns it from an alert to a lullaby, something that phone manufacturers were clearly not anticipating. The ambiguity of the cell phone ringtone comes about in the context which it is used, either alert, or lullaby.

Lastly, there is ambiguity in relationships. Although similar to ambiguity of context, ambiguity of relationships has its uniqueness. The Home Health Monitor encapsulates this well. The Home Health Monitor is a device in a home that monitors and gives feedback on the home’s emotional, physical, and spiritual health using various sensors- ie. light, temperature, door-openings, or even window condensation or hair brush tempo. Oddly enough, however, the feedback is given in horoscopes, not raw data. Punchline- although it may not be clear to the user why specific things are being tracked, they will nevertheless respond in their own unique way, depending how they interpret their own relationship with the various things being tracked.

So as a whole, we can see that tactile ambiguity pointed at users can force them into situations asking them to fill in the gaps, allowing the researchers to likely learn more about the person(s) than if they simply asked pointed questions.

Another (More Modern) Look at a Use of Cultural Probes – How Does the Team make a Kit, Start to Finish.

Reading: A Blog post by Catherine Legros on

Catherine Legros, A student and design intern at the time of writing, has boiled probing down to 4 steps, which I think are quite accurate, and easy to follow and explain.

1: Conceptualizing ideas, and finding the right activities. We can put all the possible questions or concerns we have in a comprehensive list, and see how efficiently we can make overlapping probe tasks. We don’t want to have a single unique medium for each question/task we have, as that would lead to a large kit that may not feel grounded.

2: Design and instructions. How do we organize our tasks and mediums? We by now have all of our tasks, and likely most/all of our mediums, but we need to present it in a meaningful way. For example, a to-go bag with tasks and a placemat for at-home activities. Maybe some timer to keep the user engaged. That is, some way to say ‘this today, that tomorrow, this on the last day…’ The punchline – What do we want our audience to see?

3: Execution. Making the kits. Using materials that have a friendly connotation, and final kits that have been put together in some cohesive way. This will change based on your target audience and mediums/tasks.

4: Results. We know that this is a qualitative analysis, but there are some non-obvious helpful points. The results should be analyzed as a team, as for something as abstract as this, a single set of eyes may fixate on some point, or miss another. Be ready to have your ideas change. That’s okay, and probably expected.

But How Do I, the Designer, Go About Making a Probe?


By now, we know all the buzzwords. A probe should be playful, open-ended, evocative and creative and should attempt to illicit responses that spark inspiration. But what does any of that mean to the designer?

It is impossible — and not in the spirit of the approach — to specify precisely how to create a successful Probe study.  We can offer a few tips, however, based on years of experience.

Probes are a glorified collection of tasks. It’s up to the designers to create both the tasks, and the medium(s) of response. Although these kits are bordering on an art form, we can still make some generalizations. We know that we want to keep tasks on-topic (say we’re working with internet usage, we don’t want to ask how often they go drag racing) but we also don’t want to be too direct, lest we lose a lot of creative space. Here’s some general good ideas for task inquiry:

~Ask obliquely related questions. Instead of What do you use the internet for? Try How do you gossip with friends?
~Use analogies. Try What kind of vacation spot would the internet be?
~Ask general questions. Try Computers. Blessing or a curse?
~If you want to get more specific, try indexical questions. ie. What did you use the internet for last weekend?

We can also get more specific into visual medias. Not only should we ask for things we’re curious about (obviously), we should also encourage those probed to point the cameras or whatnot in directions they wouldn’t have if we simply asked them to document their lives. While the answers to our inquiries are important, there is also a great value to unexpected details and textures uncovered.

Probes generally consist of 6-15 tasks. Having a variety of tasks can: (quote- Probetools)

  • can gather different kinds of information
  • tend to be more engaging for participants
  • play to volunteers’ strengths and interests
  • allow participants to choose which tasks to do
  • can vary in their degree of playfulness v. focus
  • allow some to fail without destroying the study

It should be noted in conclusion that, once again, this is not a scientific method. It’s a method of design, which is closer to an artistic study than a data-driven scientific one. We need to work with useful styles and themes, while staying fun and less-than-intimidating.

Cultural Probing Yields Subjective Data, and Attempting to Derive Otherwise is Likely a Waste of Time.

Reading: The Value of Uncertainty – Gaver, W, Boucher, A, Pennington, S and Walker, B from

After acquiring a general knowledge of Cultural Probes, one may believe that it is foolish to intentionally receive data that cannot be quantitatively analyzed. Groups may try to send out their own probes, asking pointed, unambiguous questions, so the responses can be directly analyzed without breaking a sweat. And after Gaver’s team’s first study using these probes in the late 90s (see Cultural Probes), several groups did.

These teams are missing the point.

Let’s open right up with the basic reasons as to why this is. If we ask pointed questions, we likely already have a good idea as to the expected answer. But what about asking exploratory questions, and then attempting to “Scientifically” analyze the results? Solely analyzing probes as an average group of data will often lead to a more defined set of data, sure, but this data will likely be mediocre in it’s reflection of the group, and completely miss unusual ideas, which are often the most inspiring. Furthermore, only looking at responses that are justifiable further constrains the imaginative possibilities of the designers.

It is understandable that this may be rather different to what many are used to. We learn starting from high school biology about quantitative data and analysis, the Scientific Method, etc. So much so that it’s likely seen as “The way to do science” to many people. If you’ve never worked in a department that focuses on qualitative analysis, it may take some convincing to see it’s value… Here’s my attempt.

Five years after Gaver’s first successful use of Cultural Probes, he is now a part of another project. This time, he and his team are pursuing new technologies for the home. They once again decided to use probes to get an understanding of given homes. (Not houses in a material sense, but homes in a personal sense to their homeowners) Some examples of items in the probes this time were – Disposable cameras, and dream recorders. The cameras, once again came with requests. Most of these requests were extremely open-ended, or arguably bordering absurdity. Ranging from “The most uncomfortable place in your home” to “Something red.” The point is not only to see how these homeowners would respond to such requests, but also give the team glimpses into parts of the home that would may have been overlooked by the homeowner. At the very least, it confided to the homeowners that the team was not coming in with a specific goal, and that this probe was in no way a facade.

The dream recorders were cheap microphone devices with only a pull tab and an LED. When a participant awoke from a dream, they pulled the tab, and had 10 seconds to describe it before the device shut off. No editing or re-doing allowed. These were, unsurprisingly, extremely provocative, but also often summarized the participants’ lives in a few words.

“What is the point of deliberately confusing our volunteers and ourselves? Most fundamentally, it is to prevent ourselves from believing that we can look into their heads… it is impossible to arrive at comfortable conclusions about our volunteers’ lives or to stand back and regard them dispassionately.”

Gaver, W.W.

The probes gave the team a feel for the individuals they were working with, as the results were intensive and multilayered, and that was the whole goal from the start. There were many interesting jumping-off points from the photos, backed up by the personalities seen in the dream recorders. It’s clear to see that these probes would not have accomplished what they did if it were not for the encouraged subjectivity from the volunteers. As Gaver describes it,

“Probology [probe-methodology] is an approach [using probes to] encourage subjective engagement, empathetic interpretation, and a pervasive sense of uncertainty as positive values for design.”

A Dive into a Specific Instance of Cultural Probe Use.

Reading: Cultural Probes by Gaver, B., Dunne, T., and Pacenti, E.

I’m going to assume that you, the reader, know the very basics of what I mean when I say “Cultural Probe.” If this isn’t the case, feel free to take a look at an earlier post of mine: What are Cultural Probes? The Basics.

A note on the context of this article (which I will refer to). In the late 1990s, a design team was assigned to, over several years, look at possible novel interaction techniques to increase the presence of elderly in three communities. Bijlmer, in Amsterdam, Netherlands; a community in Oslo, Norway; and Peccioli, a commune in Italy. The project was very open-ended, which gave the team a lot of room to do whatever they see fit.

Now that that’s out of the way, let’s dive in. First of all, let’s contextualize the design team. Starting this project, they had no data on these communities except what’s in the public domain. It was understood that these communities were diverse from the start, so there was likely not an objectively correct approach to take for any or all the communities. Given how large and diverse these communities were, the team decided on using Cultural Probes to get a better grasp on what they were really working with. The particularities of what each community lacked, to put it generally.

Given how universally general each of these proposed solutions/additions had to be to be productive to an entire community, the team obviously couldn’t take a path of design that many would consider ‘traditional,’ as this isn’t a traditional task. The team decided to take an aesthetic artist-designer approach, looking for “… cultural implications of our designs, and ways to open up new spaces for design.” Using logic-driven scientific approaches to new design will only allow one to stray but so far away from the previously set path of knowledge, whereas empirical data from informal analyses, or even stray observations found by chance, can spark new creative jumping-off points.

So they had a direction, and a method – now they needed implementation [of the probes]. Before anything was made or implemented, several challenges had to be addressed. Firstly, as we all know, next to nobody likes answering phone calls from local government officials, even though in the officials’ minds, all they are doing is trying to help the people they are calling. (If we give them the benefit of the doubt, that is. Maybe this is a poor example) This problem is exacerbated by the fact that there is likely going to be an immediate distance between the team and the community caused by the notion of having professionals flown in to help said community.

Put yourself in the shoes of someone in a community like this. These communities are inherently full of people who are at once wizened by age and likely free of the labors of work, but also still full of many people who are disabled in some way(s). Society often stereotypes the elderly as a group who are needy and fragile. Whether either or both of these are worth considering as a general fact is irrelevant when you consider that if your group is actively stereotyping a community with which you want to help, you likely aren’t going to be very productive. Taking this aura of looking-down-upon away frees up a lot of design space. It is possible to go too far the other way, however. As a team with a goal, if the interaction between you and the community is too casual, you may end up with a year of wasted time from irrelevant responses. Taking this into account, the team wanted to create probes that were interesting, open-ended, and personable while not being either too condescending or too formal.

“We wanted to lead a discussion with the groups towards unexpected ideas, but we didn’t want to dominate it.”

So let’s look at what they ended up doing. (Spoiler: It was indeed successful and productive) In making the probes, the team tried to be revealing to the groups as they asked them to be revealing to the team. There is a common theme of mutual respect that can be applied to any co-design project. The final probes contained maps, postcards, cameras,and requests for small photo albums.

Maps were bundled with small inquiries. Some of these were very straightforward (A map of the world paired with “Where have you traveled in your life?) while others were more poetic. (a map of the community, with “If Piccioli was New York” alongside of stickers of the Statue of Liberty, Times Square, or a scene of people injecting drugs) The point of these maps were to be fun, individualized, and engaging. The postcards were similar, as they all had small prompts. A reason these are attractive to the team is they have casual connotations. Unlike having a formal questionnaire dropped at your door, you can personalize your answer and mail it to the exact person on the team who will read it. The disposable cameras also had requests for each picture slot, like “Your home” or “The most attractive place around,” except that near half the slots were titled “Take a picture of anything you would like to show us.” The photo album simply requested “6-10 photos to tell us your story.”

While the Oslo group returned most all of the materials, the Bijlmer group returned a bit more than half, as they seemed less convinced about the project as a whole, and the Piccioli group returned less than half, despite being enthusiastic about the probes when the team was there in person. But all of this was still great for the team. For example – “The group in Oslo is affluent, well educated, and enthusiastic: We are proposing that they lead a community-wide conversation about social issues, publishing questions from the library that are sent for public response to… public spaces.” This is clearly something that the team would not have concluded if it were not for this open ended style of information-collection.

Although the probes were successful in their goal of familiarizing the team to the community and vice-versa, they did not directly lead to any of the following designs. Instead, they inspired the team in ways that could not have been anticipated, and it must be stated that this was the goal of the probes the entire time.

Cultural Probes are clearly useful when implemented elegantly, but it must not be understated how much thought must go into analyzing what particular challenges a set of probes must face for a given project for a given community.

Bad Design is Everywhere. Part 2, Inclusivity Boogaloo.

Reading: Don Norman on – I wrote the book on user-friendly design. What I see today horrifies me from FastCompany Business Magazine. That said, Don Norman has credentials on his own, not to be discounted through the medium of an online magazine.

Inclusive Design (design intended to help more than just one target audience), when done well, looks like a happy accident. Closed captions for those hard of hearing? Doubles perfectly in loud sports bars. Sloped curbs at intersections for the blind or elderly? Perfect for mothers with baby carriages, or travelers’ suitcases. The point is, even if you can sustain your (business, product, idea, etc.) only off of your target audience currently, we are now reaching a day and age where the percentage of elderly persons in society is growing rapidly.

“What do you mean? There are ramps everywhere, hearing aids, ‘large text modes’ on phones, and handicap accessible seats everywhere!”
I hear you, I promise. Let’s dig a little deeper so I can hopefully show you that this is a problem. Many of these anecdotes are from Norman himself (a self-proclaimed elderly man) or the people he’s around, relayed by him through this article. That said, Norman claims to still be a functioning member of society. Still able to run errands without passers-by becoming irritated at his sheer senility.

For starters, elderly people typically have less strength, dexterity (accuracy), mobility (speed), less clear vision, and fading hearing. This isn’t comprehensive or perfect for any one person, but it’s what we’ll focus on as a generalization. If we look as something as simple as groceries, it may not be obvious, but it’s possible that someone may need an over sized wrench to open a jar. Norman himself says that he and his wife choose which restaurants they prefer based on the ambient noise, as constantly loud locations will give them headaches over the course of an hour. What about something less obvious? Large-text modes on phones, in a vacuum, seem like a great idea, right? But, when a company like Apple decides to make the text on their phones a font and color that’s often neigh-invisible without being lucky and having a heavy background contrast, this mode is just a facade. (If you have an iPhone, go to your settings, really squint your eyes, and try to read the grey name of your current WiFi)

There are plenty of more examples of non-inclusive design, but I hope you get the jist. What about things designed exclusively for the elderly? That should be easy, right? Well, imagine a classic wooden walking cane from the mid 1900s. Fine, right? Well, nowadays, companies that design canes and walkers for the elderly are focused so much on function, they’ve thrown form out the window. Modern tools for the elderly are often given this aesthetic that reads ‘I’m fragile and weak and old.’ So much so that many people, against their better judgement, will forego using walkers, canes, or hearing aids for this reason.

“Are elderly people handicapped? Maybe, but so is a young, athletic parent while carrying a baby on one arm and a bag of groceries in the other,
and perhaps trying to open their car door.”

Don Norman, on high contrast or large text in phones

We must not forget that when we are designing for a non-standard audience, not only should we consider their circumstances, but also remember that customers are still human beings. We’ve proven that we can design for the disabled while being inclusive to the general public (and the vice versa less so, but we’ve still seen success). We just need to not forget the basics while we are shooting off into new technology or infrastructure at lightspeed.

Bad Design is Everywhere.

Watch (5:09): It’s not you, Bad doors are everywhere. – Vox

Think of a door, just one side. What does it look like? Does it have a knob? If so, Is the knob on the left, or right? It’s pretty consistent, as far as I’m aware, that we’ve gotten used to pulling knobs on the right, and pushing on the left. Now what about doors in the wild? Maybe in an office building, or the entrance to a hotel. These often have “push” or “pull” labeled, but why? Likely because if it wasn’t labeled, people will take an incorrect guess. This should happen around 50% of the time, assuming a regular bad door, but it’s likely you know of a door that maybe 80-90% of people guess wrong on the first try. These, we call Norman Doors. (Coined by Don Norman, Author of The Design of Everyday Things)

I’ve mentioned Participatory Design, but if we go a single step up the tree, ie. more generic/further from the user, we have something known as Human-Centered Design (HCD). We can easily understand what this means (Design centered around people), but we can dig a little deeper. Two important parts of HCD are what we can call Discoverability and Feedback. Discoverability is what causes all these door problems. Discoverability is, quite literally, the ability to discover what a user can do to this something. The problem with many doors is that, on the side meant to be pushed, they may have a large handle which can be pulled. This is obviously the opposite of what we want, so we should aim to remove this ambiguity. One of many ideal doors may have a flat metal rectangle on the side meant to be pushed, as it has a very clear Discoverability. The answer of “What are my options” is set to one- push. The other side may have a cylindrical vertical handle, as it’s much more intuitive to pull a vertical handle and push a horizontal one.

Feedback, while not entirely relevant to doors, is another important part of HCD. If you’ve ever seen Apple’s iPhone X, It has no buttons on it’s touchscreen face, but it does a decent job (so I hear) at making it obvious what each touch or set of motions does. Relatively poor Discoverability, but relatively good feedback. Now, if you’ve ever had a new car, with buttons that have terms you don’t know, and pressing them appears to do nothing at all, that is a great example of something with poor feedback.

Clearly, most of these things are not designed with malice, so why is it such a common issue? The answer is likely oversight. Underwhelming, I know, but we can avoid this, rather easily with Iterative Design. Step one: Observe people. Step two: Generate ideas or solutions. Step three: Prototype in some way. Step four: Test it! Whether that be through WoZ, or just trying the thing yourself. Just ask “Is this intuitive?” If not, then we go back to step one, with our newfound knowledge, and try again.

(Together, we can save humanity from terrible Norman Doors)

What is Participatory Design? The Basics.

Reading: Universal Methods of Design by Bella Martin and Bruce Hanington.

Participatory design is a bit broader than the other concepts previously discussed. It refers to, during most of the design process, having an active communication and engagement between the design team and the user(s). There are many ways of going about this, including, but not limited to: Cultural Probes, face-to-face activity-based co-design, or diary studies*. The point is simply that the users-to-be are directly involved in impacting the design process. How one goes about a specific type of Participatory Design (Should we use Cultural Probing? Or Diary Studies? etc.) is really up to the team. While you can mix and match methods with their own various strengths, there is likely no “Correct” answer, as this is heavily based on context. For example, in person interviews are not the greatest if your audience is across the country.

*It should be noted that Diary Studies are similar to Cultural Probes, but instead of a kit, the user interacts with a guided diary over a significant period of time.

What are Cultural Probes? The Basics.

Reading: Universal Methods of Design by Bella Martin and Bruce Hanington.

Cultural Probes are, at their core, a creative way to get a better understanding of a community, possibly in ways that the designer(s) could not have anticipated otherwise.

A Cultural Probe is something (typically a package of items) sent to people in a community with instructions of how to use the contained items, and a return address. The intent of these packages is to consider the personal context of people in said community. The goal is not to send a map and say “How exactly do you get from this point to that point?” These probes are much more open ended and exploratory in nature. A better example of something in one of these kits may be a small audio recorder with instructions of use, and have instructions – The next time you go on a walk, be it to a cafeteria, or simply your favorite place to sit and read outside, bring me with you, and talk to me! Tell me anything running through your mind. There is no such thing as unimportant, here. It is important for these probes to be informal, as we want personable data, but also thoughtful, as we want high response rates, while getting thoughtful, respectful feedback from enthusiastic users. Once again, the goal is not to get data in numbers, but to better understand the target community as a whole, in relation to our project.

What is the Wizard of Oz Technique? The Basics.

Reading: Universal Methods of Design by Bella Martin and Bruce Hanington.

The Wizard of Oz (WoZ) technique, at it’s core, is a way for testing the feasibility, functionality, and general effectiveness of a UI without requiring all the time and effort into making a prototype. This is done by creating a mock-up interface for the user to ‘use’, where someone(s) on the team (the ‘Wizard(s)’) manually implement the interface. For example, say I wanted to design a smartphone banking app, which has a few functions, and a few different pages. If I wanted to gauge how a customer was feeling about this using WoZ, I may preemptively draw out all the possible pages the user could access on small whiteboards, and when the user ‘clicks’ (in this case, points and tells me “I would like to click this”) on something, I would swap out the whiteboards and adjust any numbers, as needed. The whole point is to get a general sense from the user. ie. “Man, it’s annoying that I cant get from here to there in one click.”