Basic measures of activity and use are central to computational social science research and policy, and in a decentralized domain it is more difficult to track these fundamental measures. In this work, we illustrate this difficulty by comparing several secondary sources of this data on the federated social network Mastodon, and demonstrate that there is widespread disagreement. We briefly discuss the how ground truth becomes contextual and political in this less centralized setting, and argue that new modes of infrastructure and infrastructuring are needed in this less centralized space.

PDF