How do users process the vast amount of information available online to form opinions and make decisions? To answer this question, I analyzed individual and collective decision-making processes in a social media website, Digg. The results from behavioral experiments and computer simulations indicate that some users make decisions by following the choices of others. In particular, I found evidence of three types of following behavior – following the opinions of the whole community, following the opinions of specific individuals, and reciprocal following of pairs of users – that could contribute to the emergence of collective decisions in online communities. The results further suggest that knowing others’ opinion about a piece of information affects a user’s decisions to follow the opinion but not her perception of the information. I conclude by discussing the implications of this work for predicting trends as well as for using and designing social media websites.