My doctoral research develops a deeper understanding of a micro-macro perspective that focuses on individual motivations and their impacts on collective behaviors, especially in online communities. This dissertation includes three studies in which I use a wide variety of quantitative methods such as quasi-experiment, survival analysis, topic modeling, and agent-based modeling to help understand how inconspicuous individual motives can lead to unintended behaviors when aggregated, and the collective “tipping” processes that are involved in these consequences. Findings of a Yelp study show that the individual level status motivations distort wisdom of crowds at collective level: an over-representation of status-conferring products and an under-representation of products that are not status-worthy. Results of two Airbnb studies reveal that individual-level guest decisions are influenced by the collective aggregation of prior guest decisions. Specifically, guests’ probability of choosing different-race hosts can be increased if the hosts have a large number of endorsements from previous guests/reviewers who are from the guests’ racial groups. Yet, it is also hard for a host to gain endorsement from guests of other races due to the presence of visual and environmental cues of the majority group. An agent-based model shows that such racial segregation, reflected by the percentage of same-race guests in listings, is influenced by the racial population distributions among guests and hosts. The endorsements from previous reviewers attenuate such racial segregation. Going forward, I propose three future research directions based on the framework of the micro-macro perspective: 1) to harness individual-level motives to combat algorithmic biases at macro level, 2) to explore the effect of platform affordances such as personal contact to promote intergroup understandings, and 3) to extend micro-level theories to a macro-level scale.