10 min read

Responsible interaction design

A talk in honour of world interaction design day, suggesting contextualising history, considering power dynamics and acting (or not) with intention
NUX and IXDD 'responsible interaction design' Vicky Teinaki

This is adapated from a talk given at NUX Newcastle on 21 September 2023.

26 September is World Interaction Design Day.

First up, interaction design is IXD with an X. (ID was already taken by either industrial design or interior design, depending on your design alliances).

And it's been around for a while.

The two main organisations promoting interaction design are the Interaction Design Foundation and the Interaction Design Association (IXDA).

I signed up for the IXDA mailing list back in October 2007 (mainly as I had a gmail account even back then and so never cleared it out—yes, I know, I should do a purge as it's bad for the environment). Going through some old posts was a wonderful time capsule of what people cared about at the time, with posts including:

  • 'why has micro-blogging become so popular' (December 2007) - mentioning not only Twitter but also Pownce and Jaiku
  • 'why haven't video calls taken off' (June 2008) - cos Nokia says it makes you look ugly
  • and my favourite 'touchscreen interfaces - hype?' (September 2008)

However, aside from snarking about the mistakes of the time, it did point out some challenges that would become even more apparent in later years. A 2008 post 'US News thinks we have potential' linked to an article that mentioned "usability/user experience specialist" with extra research and engineering silos…

I've certainly noticed that the mid-late 00s seemed to be the high point of interaction design. When I look at my bookshelf for interaction design books, the ones I have include:

More recently, someone quipped that these days interaction design is the ‘red squirrel’ of user-centred design—the lovely endemic profession overrun by the ‘grey squirrels’ of newer entrant professions such as user experience design, service design, content design, and, more recently, product design. (To which I see you my 2005 bachelor in product design when product design still meant industrial design in a commercial context!)

Red squirrel with interaction design, grey squirrel with content design, user research, and product design

That said, I think that information architecture would also like a word, as that profession has been forgotten about more than interaction design—though that’s a story for another day.

Meme of kids in pool where recruiters are going after UX design, product design, service design, content design and user research. Interaction design is drowning and information architecture is a skeleton underwater

Still, I love interaction design. I love the niche of understanding the material of digital interactions and finding ways to help people do things through understanding that material.

However, as someone who notices newer generations of interaction designers coming up and learning the craft, I have noticed a few themes (which I am also going to tie to the World Interaction Design Day themes of ethics, equity and responsibility).

Responsible interaction design

I suggest that responsible interaction design means doing 3 things:

  1. Contextualising history
  2. Considering power dynamics
  3. Acting (or not) with intention

Contextualising history

All design should consider context.

Not only is 'consider context' a UK government design principle, but also was one of the principles of good design from famed industrial designer and architect Eero Saarinen “Always design a thing by considering it in its next larger context—a chair in a room, a room in a house, a house in an environment, environment in a city plan.” (Time 1956, quoting as something he learned from his father Eliel so also older ).

However, we also need to dig into the history of why things are the way that they are.

There’s an axiom known as Chesterton’s Fence that states that you should only take something down (the said fence) when you understand why it was put up in the first place.

I wrote about this in a blog post looking backwards to look forward, specifically how some services make more sense when understanding the conditions that the policy was created in, be it what was considered normal at the time (see the Industrial Injuries Disablement Benefit, whose percentage of disablement is in line with its creation back in 1948), or even what organisation the policy lived in at the time (I suspect that the Child Maintenance Service's just-before-GOV.UK look is related to it switching to the Department for Work and Pensions in 2012—just before services had to be branded as GOV.UK). This is Conway's Law, that organisations IT structures reflect the organisational structure. Marianne Belotti's Kill It With Fire talks about how Conways Law relates to IT systems. When it comes to unpicking decisions over time, the best example I have about this is Jennifer Pahlka's Recoding America, which has a whole chapter on service archaeology.

I think about this a lot in relation to interaction design patterns. Many interaction patterns can be traced to one person or at least one company—for example, the infinite scroll was created by Aza Raskin in 2006 for Humanized Reader, the pull- to-refresh action by former Apple developer Loren Brichter in 2008 for the app Tweetie (RIP Tweetie), and the hamburger menu icon by Norm Cox back in 1981 for Xerox Starr and then used in Windows 2.0. However at least 2 of these 3 interactions are considered varying levels of problematic today. (Pull-to-refresh is in my opinion still lovely, again, RIP Tweetie).

People may do a pattern because it really seems like a good idea that can then scale.

However, they may also may do it:

  • to solve a niche problem which should only be used more widely with caution -
  • because it seemed like a good idea at the time but now it isn't (unintended consequences)

A couple of design patterns that I know of in my government work are:

By understanding context, it can also then serve as a lever for understanding whether something can actually be changed. For example the Government Digital service (GDS) changed patterns such as:

just to name a few.

In fact, one of the most memorable changes I think is how GOV.UK used to let people answer forms. For any that have seen the 2019 film I, Daniel Blake, there’s a scene where he tries to fill in a form using a GOV.UK pattern of revealing extra questions on the page, he misses some questions, the form then times out and he doesn’t get his benefit. The film I believe does play the situation for dramatic effect, though it did perhaps show why by the time of the film's release GDS had in fact not used this pattern for years, having rolled out ‘one thing per page’ in 2015.

So in turn, being responsible means finding ways to talk about what is done and why. This is easier in government—there are weeknotes and blog posts and design histories—but is possible even in the private sector. Talking about the context and history helps others build on your work intentionally.

I use the phrase ‘dig’ deliberately. A 1970s Swedish book about researching your job Dig Where You Stand has finally been translated into English. It uses the example of a cement factory—where asbestos production went from a recommended one to a carcinogenic one, and is a reminder to workers that history is dictated by those that write it—and also that what is hidden can be rewritten.

To this end, it's also important to acknowledge others. For all that GDS (or at least some people from GDS) have been accused of clearing history of anything that happened before them, they did actually do a blog post of the good work happening in other departments. This is good role modelling, acknowledge the work of others.

Responsible interaction design needs to contextualise history so as to:

  • Reuse with care
  • Watch for when the conditions for ‘best practice’ change
  • Add to the context

Which leads me on to…

Considering power dynamics

Interaction design—and more generally user-centred design—has a relationship with the end-users which is sometimes under=examined. There's a lot of talk about co-design which isn't and so on. I'm not focusing on this area so much, but there is a great (if dense) book on Design Justice which is well worth looking into. I will warn you that the book is dense—I went through it in a bookclub where we did a chapter a week—but worth it. Also relevant is KA McKercher's quiz But Is It Co-Design?

What I am interested in is what it means to do interaction design as part of a team. These days, particular in agile delivery teams, interaction design works as part of a group of other specialists—developers, quality assurance testers, user researchers, content designers, business analysts—all reporting to a product manager or other key decision maker.

And this means that there is a lot of power running around.

A few years ago I wrote about the potential for power imbalance between a content designer and interaction designer that can turn it into a UX designer and copywriter.

However, more recently I was introduced to the concept of the power cube, which I think better contextualises this (and other) power imbalances.

Power cube with 3 dimensions: levels (global, national, local), spaces (closed, invited, claimed/create) and forms (visible, hidden, invisible)

The cube has some elements that are worth unpacking relating to the content designer and interaction designer example:

  • forms of power - visible is known role power, hidden informal power, and invisible power that which is available but not used. For a team, a product manager might have visible power, but an interaction designer have hidden power and a content designer invisible power
  • spaces - this can be places such as meetings and being invited to them, and even control over changing things. To this end, I think about not only meetings but even prototypes and who can change them.

This sort of thing can be tricky to be able to unpick. One useful and slightly less high risk activity I’ve used in communities of practice and also can be done with teams is one to help people understand roles, by asking 3 questions:

  1. What do you like about your role?
  2. What do people get wrong about your role?
  3. What can other roles do to help you do the best in your role?

Aside from roles, there are other dimensions to consider such as intersectionality. Introducing Intersectionality is a good start for considering this.

In practice, I'd also recommend Hack Your Bureaucracy by Marina Nitze and Paul Sinai as a practical guide to considering organisational structures.

Responsible interaction design needs to consider power dynamics so that the best people and roles contribute to the work, not just those that hold power .

Acting (or not) with intention

All of this can be enough to force designers to freeze in inaction.

The things is, design is a bit of a rollercoaster. I like the model that Nick Udall created about creativity being like this and shown below:

Onboarding - it's all in the preparation, slowing down, accelerating into the unknown, holding creative tension, finding and following the flow, illumination, elegant execution

However, sometimes designers get stuck in the middle and are unable to progress. And I'm aware that my tips about history and power could make things even worse.

To this end, I think that it's important to remember that our work is about navigating the tension between:

  • building the right thing
  • building it right
  • building it quickly

Design hypotheses help manage this, as do frameworks to help make decisions about said hypotheses. The best model that I’ve seen for this is the lean UX ‘hypothesis prioritisation’ framework (or a similar research vs design one). I've also seen more rigorous models such as RICE.

The frameworks are a reminder that we need to channel our energies and look for opportunities to just get things out and learn when they are unlikely to cause massive issues.

I think that all of the frameworks also can be distilled to 3 questions:

  1. Is this controlled enough to not cause harm?
  2. How confident are we that it won’t cause harm?
  3. If we’re wrong in our confidence, will we immediately know when we’re wrong and be able to turn it off?

If all three are yes, then it may be OK to ship and measure (focus on the measure) rather than waste effort specifically testing it.

For example:

  • Changing some terms and conditions that are already working well—ship and measure
  • Trying out completely reworded questions that have important consequences—ideally usability test, or at least do an A/B test on them

However, sometimes it's important to actually do less (another government design principle) and focus energy on other things. Tanya Snook coined the phrase of UX theatre and it's a reminder that sometimes being involved is actually worse than not being involved at all.

To this end, there's a different set of questions:

  1. Is this a bad idea?
  2. Will IXD (and wider design) have to pretend it is a good idea?
  3. Are there other ideas IXD could better add value to?

All yes: get out of the way

Responsible interaction design knows when to advocate for the value of design work, when to let things emerge, and when to step away.

Being responsible can still be cool and fun

My fun quip about being responsible is from a Dylan Moran comedy show I attended a few years back. It was a work in progress show so he needed a timekeeper. I put my hand up. “You! look responsible…" to my look of horror (I was in my 20s!) "…in a cool, fun kinda way”.

Being responsible isn’t being boring or being a blocker, it’s about using experience to avoid knowable mistakes or biases.

Book list

Yes, I know this isn't proper MLA or other citation format…

Bonus mentions: