Project: Senior Thesis
Role: UX/UI Designer
Tools: Sketch, Illustrator, Photoshop, Invision, InDesign, Paper
Online discussion is broken. With an excess of echo chambers and a distaste for truth, it is clear that we need something better. Twitter Debates is the first step towards improving the level of discourse online.
How can we get users interested in talking with people they disagree with and make it both fun and productive? The answer lies somewhere in the application of empathy, accountability, and intrinsic motivation.
Early survey results
In situations like this, where there is an opportunity for disruption in how people use digital platforms, it is appealing to create something discrete, such as bringing a new application to the table. Usually this could work if it is something that is already aligned with and improves on features that users enjoy or need such as Discord improving online gaming chats, or Slack improving business and team communication. But promoting focused and articulate debates from people that are used to complete free-reign on the way they speak online is more difficult. For one, it is mentally taxing to debate regularly in an app dedicated to that one purpose, and it is also frankly difficult to convince people to download a new app in the first place. Users are lazy, so some things need to be put right in front of their faces. While most people think that Facebook was beyond redemption, Twitter seems to be the place where most people are dedicating their time on social media.
Not only is Twitter the more popular and ubiquitous platform, it has many other qualities that make it the perfect place for conversational disruption. Namely, Twitter has built its name on being able to talk with anyone from your friends to celebrities, so Twitter as a brand already values the spreading of ideas. The regrettable part of this vision is that Twitter is easily abused. There is constant low-level arguing happening all across the board that is fueled by character limits and lack of reprimand. But while I think that a lot of horrible things happen on Twitter, its capacity for good is extravagant. The power of worldwide conversations should not be undervalued. What internet discussion is missing right now is an editor. If something can get into people’s divisive and inflammatory conversations and trim the fat of knee-jerk responses and irrational statements, then perhaps the conversation could shift. If the whole world is talking on Twitter, then it seems like a great place to start.
Information Architecture and User Flows
What counts as disruptive to conversation?
The key component of Twitter debates is to blur out messages that disrupt conversation. It's important to note that this is not censorship of content, as that would go against the core idea of spreading ideas. Instead, it censors destructive conversation techniques, so that the conversation is always moving forward in the most productive way possible. It is all part of the re-educaiton when it comes to talking with strangers online.
Debate System Testing
I needed a way to test the unique element of Twitter Debates to see if it was a valuable method of behavior regulation. The unique element being the fact that conversations on touchy subjects now have behavioral guidelines. I created a Google hangout and sourced people that resembled my personas and had them debate over divisive topics. Whenever one of them broke one of the rules I had determined, I acted as a supplement for the blurred out text (as it appears in the final design) and called them out on their behavior. At the end I sent out a survey where they could rate their discussion partner and reflect on the experience. I performed six of these tests across twelve different people to determine whether or not this feature would be effective at steering conversations in a more constructive direction.
Luckily, it absolutely was.
KNOWING THE RULES
I performed two kinds of tests. For half of them I explained that certain behaviors would result in a moderator being involved, and for the other half I told them exactly what those behaviors were. The people who did not know what the behaviors were tried to push the boundaries to see what it took to get called out. Meanwhile, the other half made a conscious effort to avoid the objectionable behaviors for the most part. I learned that it was important for the rules to be clearly laid out and accessible at any time.
Almost every time a user was called out for using inappropriate behavior, they swiftly changed their argument to be more constructive. And in all cases, an attempt was at least made to improve. Based off of these tests, the regulation of messages was an incredibly effective method of enforcing good conversation. It never once discouraged a participant from engaging entirely.
While I found out that my design was effective at controlling behavior, there’s a far cry between regulating behavior and being fun to use. Luckily, it seems as if the regulation of message tone ended up being a powerful motivation factor. What came through in these tests was that people genuinely like having to think through their arguments and be engaged in a more focused activity online. The message regulation became less of a slap on the wrist and more of an effective reward system, where instead of trying to gain points, users were trying to avoid demerits, which in turn supported higher levels of attention and concern with the task at hand. Every single participant except for one said that they wish they had more conversations like this in their survey.