Commit: Online Groups with Participation Commitments

TL;DR

It is argued that more effortful, not less effortful, membership may support many online groups, and an alternative commitment design for online groups is explored, which requires that all members commit at regular intervals to participating, as a condition of remaining in the group.

Abstract

In spite of efforts to increase participation, many online groups struggle to survive past the initial days, as members leave and activity atrophies. We argue that a main assumption of online group design -- that groups ask nothing of their members beyond lurking -- may be preventing many of these groups from sustaining a critical mass of participation. In this paper, we explore an alternative commitment design for online groups, which requires that all members commit at regular intervals to participating, as a condition of remaining in the group. We instantiate this approach in a mobile group chat platform called Commit, and perform a field study comparing commitment against a control condition of social psychological nudges with N=57 participants over three weeks. Commitment doubled the number of contributions versus the control condition, and resulted in 87% (vs. 19%) of participants remaining active by the third week. Participants reported that commitment provided safe cover for them to post even when they were nervous. Through this work, we argue that more effortful, not less effortful, membership may support many online groups.

Figures (6)

• Figure 2: A description of commitment's main effects. Committing assures participant of the intentions of other group members, sets an achievable participation goal for them, and provides a warrant for taking initiative in the conversation. Together, these factors lead to an increase in participation from group members, which over time creates a norm of participation which further encourages participation.
• Figure 3: Commitment cycle interface in Commit. A user first comes to a group (a), commits to it and can see the group (b), and then sends a message thereby fulfilling their commitment (c).
• Figure 4: Participants in both conditions received push notifications to draw them back to the app and remind them to contribute. We paired notification types across conditions so that participants received them on a similar cadence and were encouraged to take similar actions; the framing was what changed.
• Figure 5: Comparisons of member participation in groups based on two factors: (a) the number of messages sent by group members during the study period, and (b) the active status of members in the group on each day of the study. Members in the commitment condition sent more messages (around twice the median number) and were active on far more days (over two times as many, according to the median)
• Figure 6: The number of surviving active participants across the study period. Participants "survive" so long as they are not inactive a for full week; if they are inactive for that length of time, they are no longer surviving, even if they return to post after that. By the end of the study, Commit condition groups had over four times the number of "surviving" active members.