The document summarizes research on understanding how personal preferences and social influence affect sharing behaviors on social networks. It presents three key findings:
1) Personal preferences dominate individuals' decisions about what content to share, with people more likely to share items they themselves like.
2) Less than 1% of peoples' actions can typically be attributed to social influence from friends, with influence being overestimated in prior work.
3) Modeling both personal preferences and social influence may lead to a better understanding of how content spreads on networks and help improve recommendation systems and diffusion models.