Misinformation does not reduce trust in accurate search results, but warning banners may backfire.

People rely on search engines for information in critical contexts, such as public health emergencies-but what makes people trust some search results more than others? Can search engines influence people's levels of trust by controlling how information is presented? And, how does the presence of misinformation influence people's trust? Research has identified both rank and the presence of misinformation as factors impacting people's search behavior. Here, we extend these findings by measuring the effects of these factors, as well as misinformation warning banners, on the perceived trustworthiness of individual search results. We conducted three online experiments (N = 3196) using Covid-19-related queries, and found that although higher-ranked results are clicked more often, they are not more trusted. We also showed that misinformation does not damage trust in accurate results displayed below it. In contrast, while a warning about unreliable sources might decrease trust in misinformation, it significantly decreases trust in accurate information. This research alleviates some concerns about how people evaluate the credibility of information they find online, while revealing a potential backfire effect of one misinformation-prevention approach; namely, that banner warnings about source unreliability could lead to unexpected and nonoptimal outcomes in which people trust accurate information less.

Prosocial motives underlie scientific censorship by scientists: A perspective and research agenda.

Science is among humanity's greatest achievements, yet scientific censorship is rarely studied empirically. We explore the social, psychological, and institutional causes and consequences of scientific censorship (defined as actions aimed at obstructing particular scientific ideas from reaching an audience for reasons other than low scientific quality). Popular narratives suggest that scientific censorship is driven by authoritarian officials with dark motives, such as dogmatism and intolerance. Our analysis suggests that scientific censorship is often driven by scientists, who are primarily motivated by self-protection, benevolence toward peer scholars, and prosocial concerns for the well-being of human social groups. This perspective helps explain both recent findings on scientific censorship and recent changes to scientific institutions, such as the use of harm-based criteria to evaluate research. We discuss unknowns surrounding the consequences of censorship and provide recommendations for improving transparency and accountability in scientific decision-making to enable the exploration of these unknowns. The benefits of censorship may sometimes outweigh costs. However, until costs and benefits are examined empirically, scholars on opposing sides of ongoing debates are left to quarrel based on competing values, assumptions, and intuitions.

Specialization in the marketplace for ideas.

Organizations that compete for attention in the marketplace face a strategic decision: whether to target a specialized niche or diversify to reach a broader market. Previous research has extensively analyzed the specialization dilemma faced by for-profit firms. We extend the analysis to knowledge-sharing groups in the marketplace of ideas. Using data on over 1,500 technology groups collected from an online event-organizing platform over a fifteen-year period, we measure the effect of topical focus, rarity, novelty, and technical exclusivity on audience growth, retention, and sustained engagement. We find that knowledge-sharing groups benefit marginally by specializing in rare topics but not in new topics. The strongest predictor of growth and survival is whether the group is associated with technically sophisticated topics, regardless of the breadth of focus, even though technical topics are less widely accessible. We conclude that what matters is not how specialized the organization, but how the organization is specialized.

A theory of emergence: Knowledge, rewiring and innovation.

Why do social interactions linked to sharing knowledge drive the emergence of a regional technology economy? We proffer a positive theory and explanation-sketch identifying mechanisms and initial conditions in an explanation of emergence of a knowledge economy. We trace the emergence of a knowledge economy, from a small group of founding members to a regional technology economy. With the rapid influx of new people, knowledge spillover motivates technologists and entrepreneurs to reach out beyond existing contacts to explore the expanding knowledge economy and interact with new acquaintances in the search for novelty. In the course of network rewiring in knowledge clusters, individuals share knowledge and cooperate in innovation, and move to more central positions when they interact. Mirroring the trends of increased knowledge exploration and innovative activity at the individual level, new startup firms founded during this time period come to span a greater number of industry groups. Endogenous dynamics of overlapping knowledge networks lie behind the rapid morphogenesis of new regional technology economies in New York City and Los Angeles.

Political polarization of news media and influencers on Twitter in the 2016 and 2020 US presidential elections.

Social media has been transforming political communication dynamics for over a decade. Here using nearly a billion tweets, we analyse the change in Twitter's news media landscape between the 2016 and 2020 US presidential elections. Using political bias and fact-checking tools, we measure the volume of politically biased content and the number of users propagating such information. We then identify influencers-users with the greatest ability to spread news in the Twitter network. We observe that the fraction of fake and extremely biased content declined between 2016 and 2020. However, results show increasing echo chamber behaviours and latent ideological polarization across the two elections at the user and influencer levels.

Fighting bias with bias: How same-race endorsements reduce racial discrimination on Airbnb.

Recent studies have documented racial discrimination in online interactions, mirroring the historic bias observed offline. The sharing economy is especially vulnerable due to greater dependence on mutual trust in sharing a ride, residence, or date with a stranger. These services rely on user recommendations to build trust, but the effects of these peer evaluations on racial bias are only beginning to be explored. Using data from Airbnb, we examine in-group preference for same-race hosts as well as same-race recommendations. The unexpected result is that these two manifestations of racial bias are offsetting, not reinforcing. White guests largely overcame their racial bias in host selection when hosts were endorsed by previous white guests. Moreover, we found no evidence of racial bias in the affective enthusiasm of endorsements, which suggests that the preference for same-race endorsements is motivated by the race of the recommender, not the content of the recommendation.

Polarization and tipping points.

Research has documented increasing partisan division and extremist positions that are more pronounced among political elites than among voters. Attention has now begun to focus on how polarization might be attenuated. We use a general model of opinion change to see if the self-reinforcing dynamics of influence and homophily may be characterized by tipping points that make reversibility problematic. The model applies to a legislative body or other small, densely connected organization, but does not assume country-specific institutional arrangements that would obscure the identification of fundamental regularities in the phase transitions. Agents in the model have initially random locations in a multidimensional issue space consisting of membership in one of two equal-sized parties and positions on 10 issues. Agents then update their issue positions by moving closer to nearby neighbors and farther from those with whom they disagree, depending on the agents' tolerance of disagreement and strength of party identification compared to their ideological commitment to the issues. We conducted computational experiments in which we manipulated agents' tolerance for disagreement and strength of party identification. Importantly, we also introduced exogenous shocks corresponding to events that create a shared interest against a common threat (e.g., a global pandemic). Phase diagrams of political polarization reveal difficult-to-predict transitions that can be irreversible due to asymmetric hysteresis trajectories. We conclude that future empirical research needs to pay much closer attention to the identification of tipping points and the effectiveness of possible countermeasures.

Data analysis and modeling pipelines for controlled networked social science experiments.

There is large interest in networked social science experiments for understanding human behavior at-scale. Significant effort is required to perform data analytics on experimental outputs and for computational modeling of custom experiments. Moreover, experiments and modeling are often performed in a cycle, enabling iterative experimental refinement and data modeling to uncover interesting insights and to generate/refute hypotheses about social behaviors. The current practice for social analysts is to develop tailor-made computer programs and analytical scripts for experiments and modeling. This often leads to inefficiencies and duplication of effort. In this work, we propose a pipeline framework to take a significant step towards overcoming these challenges. Our contribution is to describe the design and implementation of a software system to automate many of the steps involved in analyzing social science experimental data, building models to capture the behavior of human subjects, and providing data to test hypotheses. The proposed pipeline framework consists of formal models, formal algorithms, and theoretical models as the basis for the design and implementation. We propose a formal data model, such that if an experiment can be described in terms of this model, then our pipeline software can be used to analyze data efficiently. The merits of the proposed pipeline framework is elaborated by several case studies of networked social science experiments.

Opinion cascades and the unpredictability of partisan polarization.

"Culture wars" involve the puzzling alignment of partisan identity with disparate policy positions, lifestyle choices, and personal morality. Explanations point to ideological divisions, core values, moral emotions, and cognitive hardwiring. Two "multiple worlds" experiments (n = 4581) tested an alternative explanation based on the sensitivity of opinion cascades to the initial conditions. Consistent with recent studies, partisan divisions in the influence condition were much larger than in the control group (without influence). The surprise is that bigger divisions indicate less predictability. Emergent positions adopted by Republicans and opposed by Democrats in one experimental "world" had the opposite outcome in other parallel worlds. The unpredictability suggests that what appear to be deep-rooted partisan divisions in our own world may have arisen through a tipping process that might just as easily have tipped the other way. Public awareness of this counter-intuitive possibility has the potential to encourage greater tolerance for opposing opinions.

It's not just how the game is played, it's whether you win or lose.

Growing disparities of income and wealth have prompted extensive survey research to measure the effects on public beliefs about the causes and fairness of economic inequality. However, observational data confound responses to unequal outcomes with highly correlated inequality of opportunity. This study uses a novel experiment to disentangle the effects of unequal outcomes and unequal opportunities on cognitive, normative, and affective responses. Participants were randomly assigned to positions with unequal opportunities for success. Results showed that both winners and losers were less likely to view the outcomes as fair or attributable to skill as the level of redistribution increased, but this effect of redistribution was stronger for winners. Moreover, winners were generally more likely to believe that the game was fair, even when the playing field was most heavily tilted in their favor. In short, it's not just how the game is played, it's also whether you win or lose.

Global music streaming data reveal diurnal and seasonal patterns of affective preference.

People manage emotions to cope with life's demands1,2. Previous research has identified affective patterns using self-reports3 and text analysis4,5, but these measures track the expression of affect, not affective preference for external stimuli such as music, which affects mood states and levels of emotional arousal1,6,7. We analysed a dataset of 765 million online music plays streamed by 1 million individuals in 51 countries to measure diurnal and seasonal patterns of affective preference. Findings reveal similar diurnal patterns across cultures and demographic groups. Individuals listen to more relaxing music late at night and more energetic music during normal business hours, including mid-afternoon when affective expression is lowest. However, there were differences in baselines: younger people listen to more intense music; compared with other regions, music played in Latin America is more arousing, while music in Asia is more relaxing; and compared with other chronotypes, 'night owls' (people who are habitually active or wakeful at night) listen to less-intense music. Seasonal patterns vary with distance from the equator and between Northern and Southern hemispheres and are more strongly correlated with absolute day length than with changes in day length. Taken together with previous findings on affective expression in text4, these results suggest that musical choice both shapes and reflects mood.

The strength of long-range ties in population-scale social networks.

Long-range connections that span large social networks are widely assumed to be weak, composed of sporadic and emotionally distant relationships. However, researchers historically have lacked the population-scale network data needed to verify the predicted weakness. Using data from 11 culturally diverse population-scale networks on four continents-encompassing 56 million Twitter users and 58 million mobile phone subscribers-we find that long-range ties are nearly as strong as social ties embedded within a small circle of friends. These high-bandwidth connections have important implications for diffusion and social integration.

Peer effects on adolescent smoking: Are popular teens more influential?

Previous research on adolescent cigarette adoption has focused on peer influence and the perceived status gain from smoking but has ignored the status effects on peer influence. We analyze adolescent peer effects on cigarette consumption while considering the popularity of peers. The analysis is based on a four wave panel survey representative of American high school students. We measure peers' popularity by their eigenvector centrality in high school social networks. Using lagged peers' behavior, school fixed effects, and instrumental variables to control for homophily and contextual confounds, we find that the probability of smoking the following year increases with the mean popularity of smokers, while the popularity of non-smokers has the opposite effect. These effects persist seven and fourteen years later (wave 3 and 4 of the data). In addition, the probability of smoking increases with the smoking propensity of the 20% most popular teens and decreases with the smoking propensity of the bottom 80%. The results indicate the importance of knowing not only the smoking propensity within a school but also the location of smokers within the social hierarchy.

Cultural values and cross-cultural video consumption on YouTube.

Video-sharing social media like YouTube provide access to diverse cultural products from all over the world, making it possible to test theories that the Web facilitates global cultural convergence. Drawing on a daily listing of YouTube's most popular videos across 58 countries, we investigate the consumption of popular videos in countries that differ in cultural values, language, gross domestic product, and Internet penetration rate. Although online social media facilitate global access to cultural products, we find this technological capability does not result in universal cultural convergence. Instead, consumption of popular videos in culturally different countries appears to be constrained by cultural values. Cross-cultural convergence is more advanced in cosmopolitan countries with cultural values that favor individualism and power inequality.

Measuring structural similarity in large online networks.

Structural similarity based on bipartite graphs can be used to detect meaningful communities, but the networks have been tiny compared to massive online networks. Scalability is important in applications involving tens of millions of individuals with highly skewed degree distributions. Simulation analysis holding underlying similarity constant shows that two widely used measures - Jaccard index and cosine similarity - are biased by the distribution of out-degree in web-scale networks. However, an alternative measure, the Standardized Co-incident Ratio (SCR), is unbiased. We apply SCR to members of Congress, musical artists, and professional sports teams to show how massive co-following on Twitter can be used to map meaningful affiliations among cultural entities, even in the absence of direct connections to one another. Our results show how structural similarity can be used to map cultural alignments and demonstrate the potential usefulness of social media data in the study of culture, politics, and organizations across the social and behavioral sciences.

Why Do Liberals Drink Lattes?

Popular accounts of "lifestyle politics" and "culture wars" suggest that political and ideological divisions extend also to leisure activities, consumption, aesthetic taste, and personal morality. Drawing on a total of 22,572 pairwise correlations from the General Social Survey (1972-2010), the authors provide comprehensive empirical support for the anecdotal accounts. Moreover, most ideological differences in lifestyle cannot be explained by demographic covariates alone. The authors propose a surprisingly simplesolution to the puzzle of lifestyle politics. Computational experiments show how the self-reinforcing dynamics of homophily and influence dramatically amplify even very small elective affinities between lifestyle and ideology, producing a stereotypical world of "latte liberals" and "bird-hunting conservatives" much like the one in which we live.

The mesh of civilizations in the global network of digital communication.

Conflicts fueled by popular religious mobilization have rekindled the controversy surrounding Samuel Huntington's theory of changing international alignments in the Post-Cold War era. In The Clash of Civilizations, Huntington challenged Fukuyama's "end of history" thesis that liberal democracy had emerged victorious out of Post-war ideological and economic rivalries. Based on a top-down analysis of the alignments of nation states, Huntington famously concluded that the axes of international geo-political conflicts had reverted to the ancient cultural divisions that had characterized most of human history. Until recently, however, the debate has had to rely more on polemics than empirical evidence. Moreover, Huntington made this prediction in 1993, before social media connected the world's population. Do digital communications attenuate or echo the cultural, religious, and ethnic "fault lines" posited by Huntington prior to the global diffusion of social media? We revisit Huntington's thesis using hundreds of millions of anonymized email and Twitter communications among tens of millions of worldwide users to map the global alignment of interpersonal relations. Contrary to the supposedly borderless world of cyberspace, a bottom-up analysis confirms the persistence of the eight culturally differentiated civilizations posited by Huntington, with the divisions corresponding to differences in language, religion, economic development, and spatial distance.

The social contagion of generosity.

Why do people help strangers when there is a low probability that help will be directly reciprocated or socially rewarded? A possible explanation is that these acts are contagious: those who receive or observe help from a stranger become more likely to help others. We test two mechanisms for the social contagion of generosity among strangers: generalized reciprocity (a recipient of generosity is more likely to pay it forward) and third-party influence (an observer of generous behavior is more likely to emulate it). We use an online experiment with randomized trials to test the two hypothesized mechanisms and their interaction by manipulating the extent to which participants receive and observe help. Results show that receiving help can increase the willingness to be generous towards others, but observing help can have the opposite effect, especially among those who have not received help. These results suggest that observing widespread generosity may attenuate the belief that one's own efforts are needed.

Using publicly visible social media to build detailed forecasts of civil unrest.

We demonstrate how one can generate predictions for several thousand incidents of Latin American civil unrest, often many days in advance, by surfacing informative public posts available on Twitter and Tumblr. The data mining system presented here runs daily and requires no manual intervention. Identification of informative posts is accomplished by applying multiple textual and geographic filters to a high-volume data feed consisting of tens of millions of posts per day which have been flagged as public by their authors. Predictions are built by annotating the filtered posts, typically a few dozen per day, with demographic, spatial, and temporal information. Key to our textual filters is the fact that social media posts are necessarily short, making it possible to easily infer topic by simply searching for comentions of typically unrelated terms within the same post (e.g. a future date comentioned with an unrest keyword). Additional textual filters then proceed by applying a logistic regression classifier trained to recognize accounts belonging to organizations who are likely to announce civil unrest. Geographic filtering is accomplished despite sparsely available GPS information and without relying on sophisticated natural language processing. A geocoding technique which infers non-GPS-known user locations via the locations of their GPS-known friends provides us with location estimates for 91,984,163 Twitter users at a median error of 6.65km. We show that announcements of upcoming events tend to localize within a small geographic region, allowing us to forecast event locations which are not explicitly mentioned in text. We annotate our forecasts with demographic information by searching the collected posts for demographic specific keywords generated by hand as well as with the aid of DBpedia. Our system has been in production since December 2012 and, at the time of this writing, has produced 4,771 distinct forecasts for events across ten Latin American nations. Manual examination of 2,859 posts surfaced by our method revealed that only 108 were discussing topics unrelated to civil unrest. Examination of 2,596 forecasts generated between 2013-07-01 and 2013-11-30 found 1,192 (45.9%) matched exactly the date and within a 100 km radius of a civil unrest event reported in traditional news media.