Projects

Social media connects people at an emotional level, allowing them to share their own feelings and to react to the feelings of others at an unprecedented scale and speed. Emotional connection can be a force for good, knitting people within communities that provide a shared group identity and help make sense of a chaotic world. But, it can also erode wellbeing through negative social comparisons or by trapping people within toxic echo chambers that harm mental health. Online emotions can also amplify affective polarization by exposing people to the opinions of their ideological foes, which can rapidly entrench political divides, and allow malicious actors to manipulate beliefs at scale through coordinated influence campaigns. We are mapping collective emotional dynamics at a global scale to understand the complex interplay between emotions, identity and beliefs.

Our reliance on data to fuel AI raises important questions about fairness and ethics. Social data is often heterogeneous, as it comes from a population composed of subgroups with different characteristics and behaviors. A trend in aggregate data may disappear or reverse when the data is disaggregated into its constituent subgroups. This effect, known as Simpson's paradox, often confounds models learned from data. We are developing methods to quantify biases in heterogeneous data. One approach leverages the Simpson's paradox by accounting for latent groups to learn more robust and generalizable models. We are also developing principled mathematical methods to create unbiased features for learning fair models, or use affirmative action to improve collective outcomes of interventions.

Despite long-term efforts to increase women’s representation in the scientific workforce, they continue to face barriers to advancement. We showed that women publish in less prestigious journals and receive fewer citations. The multifaceted gender disparities create a glass ceiling, an invisible barrier that fundamentally limits professional opportunities for even the best women scientists. The Covid-19 pandemic has only amplified existing gender disparities. To address these challenges, we are developing methods to audit gender biases in science. One recent example is a PNAS paper which identified gender disparities in the citations of members of the National Academy of Sciences. These differences, moreover, were strong enough to enable us to accurately predict the scholar’s gender from their citation networks. We have also developed a model of growing citation networks that explains the emergence of gender disparities in science.

Social networks shape perceptions by exposing people to the opinions of their peers. However, an individual’s perceived popularity of some trait or an opinion may be very different from its actual prevalence in the network. This perception bias arises due to the friendship paradox, which states that “your friends are more popular than you, on average.” Mind bogglingly, the stronger version of the friendship paradox is also true for the vast majority of people: “most of your friends are more popular than you.” As a result, it may seem to you that most of your friends are more successful, wealthier and have more accomplished children. They even have cuter pets! Strong friendship paradox leads to mind-bending phenomena, like the Majority Illusion, in which a rare trait can appear to be exceedingly popular within most social circles. This can bias perceptions of popularity online: we showed that some topics on Twitter appear to be several times more popular than they really are, i.e., many more people see their friends talking about it compared to how many people are actually talking about it. Explaining these paradoxes mathematically required us to define a new property of social networks -- transsortativity -- which measures the correlation of the popularity of a node’s neighbors. Friendship paradox explains why your friends seem to lead more exciting lives than you do. On the downside, it could fuel negative social comparisons that are detrimental to mental health and wellbeing.

The many decisions people make about what to pay attention to online shape social interactions and the spread of information on social media. Due to the constraints of available time and cognitive resources, the ease of discovery strongly impacts how people allocate their attention to social media content. As a consequence, the position of information in an individual’s social feed determines whether it will be seen, and the likelihood that it will be shared with followers. Accounting for these cognitive limits explains puzzling empirical observations: (i) information generally fails to “go viral” (ii) highly connected people are less likely to re-share information due to their higher information load. In addition, we often observe “performance deterioration”, wherein performance quality declines over the course of an online session, demonstrating that “attention is a finite resource.” We are studying the interplay between human cognitive limits, content and network structure to better understand and predict online interactions.