Thus, given the vagaries of social science and the challenge of figuring out how to answer the causal question (without randomly assigning 300 children to avoid social media until they turn 18 and have their narcissism measured), I think it is best to wait for more data.