Slowing it Down: Towards Facilitating Mindfulness in Heated Online Conversations
This article summarizes the work I will present at the 26th ACM Conference on Computer-Supported Cooperative Work (CSCW 2023). Read the full paper here.
In many online spaces, people discuss polarizing and contentious topics with one another. It’s important to have these discussions to promote understanding and awareness across ideological gaps, and create inclusion in an increasingly divided society. Social media platforms offer a space to have these conversations by bringing people together from diverse belief systems. Unfortunately, these discussions can quickly become distressing, unproductive, and harmful.
In this work, we look at how to promote mindfulness during contentious online conversations using the design features of communication platforms. Our approach involves altering the conversational timing of synchronous, one-on-one, text-based discussions. We ask: what happens to a contentious online debate when users can only send one message every two minutes?
To explore the effects of slowing down polarizing conversations, we conducted a controlled experiment with 40 participants. During recruitment, we gathered their opinions on a variety of topics and paired them with partners who held opposing viewpoints. All 20 “disagreeing pairs” were asked to complete a negotiation task in a private chatroom on Discord. Their task was to attempt to come to an agreement on how to hypothetically allocate funding to three different sociopolitical movements. Half of the pairs had “Slow Mode” enabled, meaning they could send one message every two minutes. The other half of the pairs had no temporal restrictions and could send messages freely throughout their conversations. We analyzed all conversations across both groups, administered surveys after each conversation, and interviewed a cross-section of participants from both groups.
Did participants behave differently when their conversations were slowed? First, we noticed that the structure of participants’ conversations changed drastically. Participants in the Slow Mode Group wrote much longer messages than those in the Control Group. Their messages were more detailed, addressing multiple points with a more efficient tone. In an interview, a Slow Mode participant explained their experience with the delay:
“Once I hit that enter button, it was two minutes before I could say anything else. To keep the conversation going, you want to say meaningful, insightful, productive things.”
It seemed that the delay caused each individual message to hold more value, leading participants to write with more intention. The Slow Mode Group was also more focused on the common goal of their conversations, making sure each message was useful toward coming to an agreement and finishing the task. Whereas, in the Control Group, participants sent many short messages that were not always related to the task.
Another finding stood out during analysis: participants in the Slow Mode Group exceeded the 2-minute delay, taking an average of four minutes to send each message. In the Control Group, the average time between sending messages was just over one minute. This means that Slow Mode participants were not simply waiting to click “send” as soon as the two minutes elapsed. Instead, the delay acted like a nudge, encouraging participants to put even more time into each message before sending. In an interview, a Slow Mode participant described their thought process:
“[SlowMode] really forced me to stop and think, ‘Okay, what’s coming my way, and what am I going to say’ […], instead of just blurting out.”
Participants in the Slow Mode group appreciated the increased task-focus and thoughtfulness brought on by Slow Mode. However, they were also frustrated by the delay because it interrupted their natural conversational flow, and it limited their ability to build rapport with their partner.
This sort of dichotomy is seen elsewhere in research about design frictions and micro-boundaries. Sometimes, it is important to interrupt a seamless interaction to actually make the interaction effortful and seam-ful. This may annoy a user, but it can help jolt them out of an automatic, reactive, and mindless way of using technology, into a more mindful interaction style.
Admittedly, none of the conversations in this study reached the apex of toxicity that is of most concern in social media today. This is likely because participants knew in advance that they would have a potentially contentious conversation, and partners shared the common goal of reaching an agreement and completing the task.
Still, it is worth exploring a “slowing it down” type of intervention for communication that has the potential to become toxic, due to the nature of the situation. For example, future research would be fruitful in high-stakes contexts like sensitive legal negotiations, or interactions between separated co-parents or partners in a hostile relationship. Individuals in these situations have a specific and important purpose for communicating with one another, but their communication may be very tense or heated. This is when slowing down communication may be most beneficial to keep the conversation productive and civil.
