Predicting length of membership in online community "everything2" using feedback
Chandan Kumar Sarkar, Donghee Yvette Wohn, Cliff A.C. Lampe · 2012
In this paper, we examine how specific features of participation and feedback can predict the length of membership within a user generated content based online community called 'Everything2'. Examining almost 10 years of server data we found that not all feedback is the same: feedback on the user's initial contribution was the strongest factor explaining membership length. Receiving one negative initial feedback did not significantly affect membership, but sequential negative initial feedback decreased the likelihood of staying longer on the site.