Study of Studies

Million-Dollar Titles: The Story of Clickbait

When you hear the word “clickbait,” the experiences that come to mind probably aren’t good ones. Whether it’s getting rick-rolled for the thousandth time or having been let down by a misleading title on Youtube, people often feel frustration and anger toward digital media outlets that design headlines and thumbnails for the purpose of maximizing clicks. But how malicious is clickbait, really, and could there be a reason why journalists and content creators continue to use it?

The answer has to do with how news organizations operate within the contemporary media environment. Nowadays, traditional news organizations and newer, metrics-based media outlets compete with each other for public readership and engagement – researchers have called this the era of “Clickbait Media” (Munger 2019). However, this is not a new phenomenon. Tabloid headlines have been a feature of news publications since the early 1900s, under the name of “yellow journalism.”

In both yellow journalism and clickbait media, content creators use language and images with emotional appeals to target a younger audience (Zannettou 2019). Compared to previous forms of tabloid reporting, clickbait is distinct in that it appeals to relatively positive responses, such as surprise (“She said yes!”) and intrigue (“… You won’t believe this happened next.”) (Chakraborty 2017). Social media also brings an interactive aspect to news publications, where creators would encourage viewers to share the work using tags, mentions, retweets and the like. People will attach value to articles based on population indicators, in what is known as the bandwagon effect (Jiang 2019). Whether people admit it or not, these strategies seem to work – clickbait articles can usually reach a wider audience than their non-clickbait counterparts (Chakraborty 2017).

Skeptics may be quick to suggest that overly widespread use of clickbait could exacerbate the effects of partisan media, driving people to look for sources that reinforce their opinions and separate themselves from those who don’t share their ideology. In the current moment, there is no empirical evidence for this. Although members of society may adopt more extreme ideologies due to other factors, researchers have found there to be no relationship between clickbait consumption and polarization (Munger 2019). Rather, the general consensus in contemporary media effects research is that an individual’s response to news is dependent upon preexisting beliefs and theoretical frameworks (Schefele and Tewksbury 2007). In other words, clickbait does not seem to be encouraging social divisions in and of itself, because we are less persuadable than we think.

But what if clickbait enables the spread of false information that is manipulative? In 2016, many became concerned by “fake news” when misinformation circulating on Facebook affected the results of the presidential election. To prevent further spread of misinformation, researchers are in the process of developing machine learning techniques for programs to automatically evaluate online content for credibility. Models that evaluate the overall argumentative structure of an article seem to be most effective, more so than models for authenticating users and analyzing content (Meel and Vishwakarma 2019).

Identifying clickbait is another story. Researchers have been able to systematically identify clickbait on Youtube by determining whether the content of a video matches its title, and by assessing chosen tags for the video (if one of them is “sexi”, there is a 95% chance it’s clickbait) (Zannettou 2018). Other researchers have detected clickbait through audience reactions: comments to clickbait videos are usually shorter, centered around singular threads, and discuss the content of the video itself less (Shang 2019).      

What clickbait will look like in the future isn’t clear, as social media platforms continue to define what is acceptable on their platforms and what isn’t. Upworthy, the clickbait-driven news outlet, had more site visits than the New York Times and the Washington Post at its peak in mid-2013 (Munger 2019). The company’s subsequent decline stemmed from a change in algorithms that downweighted its content. At the same time, mainstream news outlets have started to adopt clickbait strategies to expand the reach of their articles. Some academics are also hopeful that clickbait strategies of popularity and audience-building can lead to more people tuning into public affairs and increase political engagement among younger voters (Chakraborty 2017).

The next time you fall for clickbait, you might want to think about why the author took such lengths to hook you in. Perhaps that message is meant to be shared.

The Studies

Chakraborty, Abhijnan et. al. 2017. “Tabloids in the Era of Social Media? Understanding the Production and Consumption of Clickbaits in Twitter.” Proc. ACM Hum.- Interact.1(CSCW): Article 30.

Jiang, Tingting et. al. 2019. “What prompts users to click on news headlines? Evidence from

unobtrusive data analysis.” Aslib Journal of Information Management 72 (1): 49-66.

Meel, Priyanka, and Dinesh Kumar Vishwakarma. 2019. “Fake News, Rumor, Information Pollution in Social Media and Web: A Contemporary Survey of State-of-the-arts, Challenges and Opportunities.” Expert Systems with Applications 153 (1): Article 112986.

Munger, Kevin. 2019. “All the News That’s Fit to Click: The Economics of Clickbait Media.” Political Communication37: 376-397.

Munger, Kevin et. al. 2020. “The (Null) Effects of Clickbait Headlines on Polarization, Trust, and Learning.” Public Opinion Quarterly 84 (1): 49-73.

Scheufele, Dietram A. and Daivd Tewksbury. 2007. “Framing, Agenda Setting, and Priming: The Evolution of Three Media Effects Models.” Journal of Communication 57: pp. 9-20.

Shang, Lanyu et. al. 2019. “Towards reliable online clickbait video detection: A content-agnostic approach.” Knowledge-Based Systems 182.

The Editors. 2015. “A Growth Spurt.” A Thing About Words. Retrieved from 2015/05/a-growth-spurt/.

Zannettou, Savvas et. al. 2018. “The Good, the Bad and the Bait: Detecting and Characterizing Clickbait on YouTube.” IEEE Security and Privacy Workshops (SPW): pp. 63-69.

Zannettou, Savvas et. al. 2019. “The Web of False Information: Rumors, Fake News, Hoaxes, Clickbait, and Various Other Shenanigans.” ACM Journal of Data and Information Quality 11(3): Article 10.

Study of Studies

What Research says about the Effects of Contact with Immigrants

People possess all kinds of stereotypes and false information about strangers, especially about those outside our “circles.” Until we make personal contact with these outsiders, misperceptions about them are likely to persist. For one, I remember when Americans were surprised when they first learned that not every Chinese person knows Kung Fu.

Just like you’d expect, social scientists have long found that contact with out-group members can reduce prejudices about them from in-group members. Contact generates new information that helps in-group members correct misperceptions, increase empathy, and develop emotional ties (Allport, 1954).

This idea leads to the question: could an increased contact with immigrants reduce xenophobia and stereotypes among Americans? We do have some evidence to support this hypothesis. If one can recall the whole stunt of the caravan “invasion” in 2018, Americans who meet more immigrants are less likely to believe in it (Murray, 2018). More rigorous academic research, however, presents mixed results even when we limit our scope to those studies conducted in the U.S. While some research suggests that contact with immigrants can reduce prejudice, increase the perceived value of immigrants, and yield more support for inclusionary policies, others claim contact actually leads to stronger exclusionary reactions and feelings of threat.

Let’s first consider a study that asks this question: if you commute by train from work to home and suddenly one day an unusually large number of immigrants showed up, would you expect to instantly feel warmer towards them? Enos (2014) conducted such an experiment where he randomly assigned pairs of Spanish-speaking confederates to visit train stations in Boston, a homogeneously Anglo community, for two weeks. Enos found that repeated intergroup contact led to more exclusionary attitudes toward immigrants. Why didn’t contact have positive effects here?

As it turns out, the effects of contact are not universal – who and how condition the effects. One such essential condition is the “friendship potential”: the contact process much present real opportunities for immigrants and natives to become friends, and that typically requires interactions across times and different social contexts (Pettigrew 1998).

In Enos’ study, the demographic change he created only represents a superficial form of contact – seeing more immigrants in the community. Without the potential to develop friendship, the mere presence of outgroups is more likely to induce threats than reducing prejudices. Enos even suspected that repeated exposure can mitigate the initial negative reactions. Surveys of natives sometimes only ask for “casual contact,” which is unlikely to reduce the perception of threat (Gravelle 2016).

Other research that uses a different method also confirmed the “friendship potential” as an essential condition to generate positive effects from contact. Ellison and his colleagues (2011) asked how different aspects of contact with Latinos affect attitudes toward the U.S. Latinos and immigration restrictions. They found that the most consistent predictor of positive views of Latino immigrants and immigration policies is having Latino friends, followed by relatives. 

To become friends, ideally, there should be few cultural and language barriers. A 2012 mixed-method study (Newman, Hartman, and Taber 2012) showed that white Americans who came in contact with non-English speaking immigrants enhanced their perceptions of immigration as threats and expressed more support for exclusionary immigration policies. One interpretation is that natives and immigrants who have higher language skills and cultural exposures are more likely to develop positive viewpoints of one another.

Perhaps unsurprisingly, partisanship also plays a large role in conditioning the effect of contact in the United States. Survey studies found that contact only has threat-reducing effects among voters on the left (Homola and Tavits 2018). Democratic voters who are predisposed to political values like equality and tolerance are more likely to positively update their views of immigrants than Republican voters who tend to oppose social changes. In addition, for positive effects of contact to sustain, in-party members cannot provide contrary messages (Pearson-Merkowitz, Filindra, and Dyck 2016). When individuals are interpreting policy information, partisanship serves as a stronger heuristic that cancels out the positive effect of intergroup contact.

Besides partisanship, other personal traits also play a moderating role. People who have histories of positive contact with immigrants are more likely to report positive ones in the future. We also know prejudiced people less likely to engage in intergroup contact, though it’s unclear if they would report lower levels of both positive and negative contact (Kotzur, Tropp, and Wagner 2018).

The big lesson is that while more contact with immigrants can yield both more positive and negative effects, the positive ones are more common (Kotzur, Tropp, and Wagner 2018). Studies show that contact frequency consistently predicts a higher willingness for U.S.-born to welcome immigrants. Meta-analyses of over 500 studies concluded that intergroup contact typically reduces prejudice (Pettigrew and Tropp 2006; Pettigrew and Tropp 2008). In short, definitely reach out to immigrants around you, but remember that the effects depend on the quality of the interaction. 


Allport, G. W. 1954. The Nature of Prejudice. Cambridge, MA: Perseus Books.

Ellison, Christopher G, Heeju Shin, and David L. Leal. 2011. “The Contact Hypothesis and Attitudes Toward Latinos in the United States*.” Social Science Quarterly 92(4): 938–58.

Enos, Ryan D. 2014. “Causal Effect of Intergroup Contact on Exclusionary Attitudes.” Proceedings of the National Academy of Sciences 111(10): 3699–3704.

Gravelle, Timothy B. 2016. “Party Identification, Contact, Contexts, and Public Attitudes toward Illegal Immigration.” Public Opinion Quarterly 80(1): 1–25.

Kotzur, Patrick F, Linda R. Tropp, and Ulrich Wagner. 2018. “Welcoming the Unwelcome: How Contact Shapes Contexts of Reception for New Immigrants in Germany and the United States.” Journal of Social Issues 74(4): 812–32.

Murray, Patrick. 2018. National: Public Divided on Whether Migrant Caravan Poses a Threat. Monmouth University Polling Institute.

Newman, Benjamin J, Todd K. Hartman, and Charles S. Taber. 2012. “Foreign Language Exposure, Cultural Threat, and Opposition to Immigration.” Political Psychology 33(5).

Pearson-Merkowitz, Shanna, Alexandra Filindra, and Joshua J. Dyck. 2016. “When Partisans and Minorities Interact: Interpersonal Contact, Partisanship, and Public Opinion Preferences on Immigration Policy.” Social Science Quarterly 97(2): 311–24.

Pettigrew, Thomas F. 1998. “Intergroup Contact Theory.” Annual Reviews of Psychology 49: 65–85.

Pettigrew, Thomas F. 2006. “A Meta-Analytic Test of Intergroup Contact Theory.” Journal of Personality and Social Psychology 90(5): 751–83.

Pettigrew, Thomas F, and Linda R. Tropp. 2008. “How Does Intergroup Contact Reduce Prejudice? Meta-Analytic Tests of Three Mediators.” European Journal of Social Psychology 38: 922–34.

Tavits, Margit, and Jonathan Homola. 2018. “Contact Reduces Immigration-Related Fears for Leftist but Not for Rightist Voters.” Comparative Political Studies 51(13): 1789–1820.

Tropp, Linda R, Dina G. Okamoto, Helen B. Marrow, and Michael Jones-Correa. 2018. “How Contact Experiences Shape Welcoming: Perspectives from U.S.-Born and Immigrant Groups.” Social Psychology Quarterly 81(1): 23–47.

Study of Studies

Coercion-Resistant Passwords: The End of Compelled Decryption?

2016 was a year of massive data security breaches. Targets included Yahoo, Verizon, and the Clinton campaign, among others. Amid the alarm, cybersecurity professionals are declaring the simple numerical password — well, passé. A survey of 600 security experts sponsored by mobile ID provider Telesign revealed that 69% of the respondents did not think passwords provide enough security.

Data Breaches
Source: Information is Beautiful

In response to the mounting concern over data protection, 72% of businesses plan to phase out passwords entirely by 2025, foregoing them in favor of more secure alternatives such as biometric scanners and two-factor authentication.

But while biometrics and 2FA are certainly more difficult to crack than simple numerical combinations they are still not secure enough for some. They still possess a lingering vulnerability: they cannot detect when a user is being coerced to authenticate, either by typing in a code or placing their finger on a biometric scanner.

To address this vulnerability, researchers have started proposing designs for coercion-resistant passwords. A team from Stanford University has devised a proof-of-concept model for a system that issues subliminal passwords to users based on an individual 30-40 minute “training session” resembling a video game. Users of the system can never reveal their passwords — even under threat of coercion — because they simply do not know them. Additionally, researchers at California State Polytechnic University Pomona have developed an authentication system that authenticates based on user’s subconscious physiological responses to music samples. If the system detects duress, possibly resulting from the threat of coercion, during the sample designed to induce subconscious relaxation, it refuses to grant access.

So far, experts intend these coercion-resistant password systems for government or commercial use — to restrict access to top-secret government data or sensitive aggregate financial information, for example. In this respect, they might serve to considerably enhance government and corporate data security protocols.

But there is little reason to assume that these emerging technologies, once introduced, will not be subject to function creep, particularly amid the growing public demand for more secure personal data protection. Much like encryption technology, coercion-resistant password systems may eventually become available to citizens to incorporate into their personal data protection protocols.

The proliferation of coercion-resistant password systems would pose a significant challenge for law enforcement as well as for the courts that has thus far been overlooked: the coercion-resistant password would effectively nullify government-requested compelled decryption.

Courts at the district, state and federal levels, are encountering compelled decryption cases with increasing frequency, because private citizens have been using encryption technology to protect their personal electronic devices more frequently. As of now, encrypted devices are virtually impossible to access without knowing the encryption password. As a result, when law enforcement now seizes an electronic device during an investigation and finds itself unable to access the device’s contents because of an encryption lock, they typically appeal to the courts to compel the suspect to unlock the device.

The courts do not always compel decryption. Some courts have sided with the government and compelled the suspect to decrypt, while others have held that compelled decryption violates the suspect’s Fifth Amendment privilege against self-incrimination. But nonetheless, law enforcement has thus far had the recourse of appealing to the courts to compel a suspect to decrypt a device when they have been unable to access the device’s contents through independent investigative means. Coercion-resistant password systems, however, would void court-ordered compelled decryption. With the Stanford design, the user could not consciously recall the password required to unlock the device. With the Cal Poly system, the user would not be able control his or her subconscious physiological response under the duress of being compelled to decrypt, and the system would deny access because of detected coercion.

What is the proper balance between law enforcement’s investigatory privilege and the individual’s right to privacy and privilege against self-incrimination? How will novel technologies, such as next-generation encryption technology and coercion-resistant password systems, shift that balance?



“Beyond the Password: The Future of Account Security,” Lawless Research Report, sponsored by Telesign, 2016.

Gareth Morgan, “Scientists Mimic Guitar Hero to Create Subliminal Passwords for Coercion-Proof Security”, 2012.

Max Wolotsky, Mohammad Husain, and Elisha Choe. “Chill-Pass: Using Neuro-Physiological Responses to Chill Music to Defeat Coercion Attacks.” arXiv preprint arXiv:1605.01072 (2016).

Val Van Brocklin, “4 Court Cases on Decryption and the Fifth Amendment,”

Study of Studies

Teams in the Modern American Political Arena

In discussions about the causes and effects of American political polarization, the idea of “sorting,” an increasing correlation over time between characteristics in a group, is likely to play a primary role. In the context of polarization, these characteristics are associated with politics; for example, more Democrats are identifying as liberals over time (and vice versa), and more political party members are relocating to the same places as other party members over time.

Political scientists debate whether polarization is happening at all, or whether it consists entirely of partisan-ideological sorting within both groups of elected officials and the mass electorate [1]. This argument highlights the difference between issue distance, the ideological distance of mean issue positions between political groups, and issue consistency, the variation of issue positions within groups. If there is very little variation in the positions held by group members, that group is considered highly sorted.

Over the last several decades, issue distance between Democrats and Republicans- the two largest political groups- has increased relatively little, while issue consistency in both these groups has increased substantially [2]. Consistency in living areas has also increased in the last twenty years or so, and liberal Democrats are becoming more likely to live near other liberal Democrats, etc [3]. This trend might be surprising to some, as Democrats and Republicans are traditionally portrayed as diametrically-opposed forces. But if polarization is only defined as the difference between group ideological or issue positions, it doesn’t appear to have been occurring in the past several decades in America. Polarization may instead simply be a misnomer for “political sorting.”

However, many people would instinctively disagree with this conclusion. They feel rising tension in our modern political environment and likely attribute it to differences (read: disagreements) between Democrats and Republicans, liberals and conservatives. Since differences in issue positions can’t explain this tension, some political scientists focus not on issue consistency or other types of sorting to describe polarization but rather on the behavioral effects of that sorting. Sorting of politically associated identities — party, ideology, issue positions, living area, demographics — in individuals has been known to increase their affective polarization, or their active dislike for opposing political groups (and reciprocal loyalty to their own group) [4]. In fact, even the perception of demographic sorting in political parties may increase affective polarization in those who hold that perception [5].

These results hint at how sorting is related to the more tangible aspects of polarization, namely anger, distrust, and stubbornness. As more Democrats become liberal (this does not mean the same thing as “Democrats become more liberal”), more Republicans become conservative. Similarly, if more liberal Democrats move to one place while conservative Republicans move to another, or more people of different races or religions join the same party as other people of their race or religion, people begin to get more angry at the political group that is not their own, and they become more willing to fight for their own group. This does not mean that they hold very different beliefs about the world than members of the other group; it just means that they hold different, heavily sorted group identities. Given these differences in sorted group identities, we shouldn’t be surprised that tension fills our political arena

Study of Studies

Online and Offline: How Your Friends Affect Your Politics

In an ideal democracy, we would talk about politics all the time. Imagine your friends and co-workers chatting about the latest political controversies, openly disagreeing, but still civilly exchanging ideas and arguments. Rather than debate, with both sides trying to prove each other wrong, it would be a deliberative effort, in search of common ground. Someone might change their mind occasionally after weighing the evidence. Just to be clear, this does not mean the “Can you believe Donald Trump is still ahead in the polls?” kind of political talk, so much as “I believe in policy X because…” or “I support candidate Y because…” If you’re having trouble picturing this, it’s probably because this is not how political deliberation usually plays out, whether in person or online.  Political deliberation has been studied both face-to-face and over social media, and the results diverge unexpectedly.

Besides the common belief that politics is a taboo subject, research has shown that most of us are inclined to discuss politics with people with whom we agree. In fact, we self-select into social circles composed of people who already share our opinions, hindering our exposure to different (cross-cutting) views[1]. However, some scholars think that this is an exaggerated problem. The majority of people know at least one person in their social network with whom they talk politics that holds views different than their own. People in networks with disagreement generally hold less polarized viewpoints[2], and exposure to disagreement breeds political tolerance[3]. Unfortunately, disagreement also makes people less enthusiastic about politics, but the evidence is mixed on whether or not this actually affects participation, such as voting[4].

There’s some debate about whether political discussion has the same effects online, particularly on social media. Research shows that people are most likely to encounter cross-cutting viewpoints in online settings that aren’t centered around politics, but where politics keeps coming up anyway[5] (like Facebook). In fact, there is a correlation between social media use and exposure to differing viewpoints[6]. However, unlike conversations in person, disagreement on social media has been found to result in increased polarization[7] and increased political participation online, such as sharing political content[8]. It is unclear why this difference exists. Are we more biased about the information we see online? Is online communication simply too impersonal for political persuasion? There are many unanswered questions about the role of social media in politics, making it a promising subject area for future research.


[1] Mutz, Diana C. “Cross-cutting Social Networks: Testing Democratic Theory in Practice.” American Political Science Review 96.01 (2002): 111-126. Web.

[2] Huckfeldt, Robert, Jeanette Morehouse Mendez, and Tracy Osborn. “Disagreement, Ambivalence, and Engagement: The Political Consequences of Heterogeneous Networks.” Political Psychology 25.1 (2004): 65-95. Web.

[3] Mutz (2002)

[4] Huckfeldt et al., (2002)

[5] Wojcieszak, Magdalena E., and Diana C. Mutz. “Online Groups and Political Discourse: Do Online Discussion Spaces Facilitate Exposure to Political Disagreement?” Journal of Communication 59.1 (2009): 40-56. Web.

[6] Kim, Yonghwan. “The Contribution of Social Network Sites to Exposure to Political Difference: The Relationships among SNSs, Online Political Messaging, and Exposure to Cross-cutting Perspectives.” Computers in Human Behavior 27.2 (2011): 971-77. Web.

[7] Lee, Jae Kook, Jihyang Choi, Cheonsoo Kim, and Yonghwan Kim. “Social Media, Network Heterogeneity, and Opinion Polarization.” Journal of Communication 64.4 (2014): 702-22. Web.

[8] Kim, Yonghwan, and Hsuan-Ting Chen. “Social Media and Online Political Participation: The Mediating Role of Exposure to Cross-cutting and Like-minded Perspectives.” Telematics and Informatics 33.2 (2016): 320-30. Web.