In July 2017, I participated in the workshop Designing for Disagreement at Copenhagen Institute of Interaction Design. I was drawn to this workshop because we increasingly live online and we disagree now more than ever. My instructors delivered lectures on behavioral design, psychology, and ethics in the mornings. In the afternoons my team and I used service prototyping techniques to test new ideas for civic interaction.
The workshop structure:
Online conversations have skyrocketed and the expectations of them have changed. Facebook, one of the main places where discussion, sharing, and consumption happens online, now has 1.7+ billion users. We live online so fully, that even when we disagree, we don't sign off. Instead we seek out places where we do agree, like closed threads and private groups.
Personalized digital feeds are so tailored and specific, our feeds are basically closed groups of our making. We block, unfriend or unfollow whoever we choose. We're always going to disagree, but what if encouraged users to disagree in more productive ways?
My team's goal for the 5-day workshop was to construct a hypothesis statement and test it with as many people as possible. Here's a peek into our process:
We pursued an idea of asking bullies to stop and think before posting. The idea we chose to test is a NLP filter that detects potentially abusive language and asks teens if they want to post what they've written. They can select 'yes' and post, but if abusive language has been detected they have to wait 5 minutes. We learned from behavioral science that people respond better to an opportunity to change their own decision rather than a punishment.
In testing, we received a range of responses. Some thought that it would be successful because it would give people a chance to stop and consider their actions. Others thought it would only delay the inevitable, and perhaps even anger online bullies and result in more bullying posts. We used paper prototypes to talk to a half dozen teenagers.
"If you use a 5-minute timer, decent people will not post and the number of bullying and harassing postings will increase relatively, won't they? That kind of guy will post anyhow."
"This will give people a chance to stop and think, so it could stop some bullying. I think people would be surprised to see this."
On the last day of the program we all shared our work with students in other workshops, and I definitely found some time to enjoy Denmark. I truly valued the lectures on behavioral science and psychology, and how much my teachers encouraged us to challenge and be rigorous about the research process, even though we only had a week. Sometimes in design, it feels easy to glom onto a research insight if you're excited about it, and this workshop helped me learn how to step back and be honest about what we were really hearing during field interviews.
My biggest takeaways
Don't take it a prompt at face value. When designing to solve a complex problem, look for initial biases in how problems are framed, like word choice, tone, body language.
Think about supporting an idea rather than loving it. Adopting this mindset helped me from getting excited about one idea then just assuming we should carry on with it.
Copenhagen rocks. Danish people are incredibly nice and in my experience, willing to talk to a stranger for 10 minutes 100% of the time, even if that stranger is asking about a potentially divisive topic.
Be consistent on word choice. Vocabulary around this topic varies widely. We used the words bullying, harassment and trolling interchangeably. Some people had not heard of the term trolling and thought we were talking about the 'live under a bridge' kind of trolls. If continuing with this research, we would need to devise a more rational approach to word choice to ensure conflating terms didn't influence data.
I am assuming...Keep an open record of assumptions, like an accountability record. This will help separate what is rooted in user research and what you think you should do. Beware logical fallacies in the design process. For example, it's easy to allow the bandwagon fallacy to drive ideas: one person says they does this, so it must be how all users who feel this way would act. Or the genetic fallacy, which would lead to giving certain user research results more credibility than others based on who said them (e.g. not taking a young user seriously). Check out the logical fallacy chart: