‘They can’t fool me,but they can fool the others!’ Third person effect and fake news detection

Abstract

The aftermath of the 2016 US Presidential Elections and the Brexit campaign in Europe have opened the floor to heated debates about fake news and the dangers that these phenomena pose to elections and to democracy, in general. Despite a growing body of scholarly literature on fake news and its close relatives misinformation, disinformation or, more encompassing, communication and information disorders, few studies have so far attempted to empirically account for the effects that fake news might have, especially with respect to what communication scholars call the third person effect. This study aims to provide empirical evidence for the third person effect in the case of people’s self-perceived ability to detect fake news and of their perception of others’ ability to detect it. Based on a survey run in August 2018 and comprising a national, diverse sample of Romanian adults (N = 813), this research reveals that there is a significant third person effect regarding people’s self-reported ability to spot fake news and that this effect is stronger when people compare their fake news detection literacy to that of distant others than to that close others. Furthermore, this study shows that the most important predictors of third person effect related to fake news detection are education, income, interest in politics, Facebook dependency and confirmation bias, with age being a non-significant predictor.

Keywords

Distant and close others fake news predictors third person effect

Introduction

There is an awful lot of hype around ‘fake news’ these days. The rapid development of social media platforms and the boundary-free, loose communication that they facilitate has contributed to the spread of news and information that is intended to mislead readers and, thus, to affect their understanding of the social, economic and political reality. Widely publicized incidents involving the use of social networks, such as Facebook, to intentionally feed people false information that might influence their vote in the elections have fueled intensive debate about the future of democracy and elections in a ‘post-truth’ world characterized by gross oversimplification, lies, fake news, half-truths, trivia (Ball, 2017). While acknowledging the negative impact that the spread of misinformation and fake news has on democracy, both social media representatives (e.g. Facebook, Twitter, Google) and public institutions (e.g. The European Commission) have taken steps to advance technological developments (e.g. Facebook has recently announced the establishment of its election War Room in which software updates and redesign are tested to root out disinformation, monitor fake news and delete fake accounts) and to draft policy initiatives to counter fake news (e.g. EC’s High-Level Expert Group on Fake News and Online Disinformation). However, despite the private and public mobilization of resources to combat fake news, a question remains rather unanswered: to what extent do people think that they can distinguish between ‘real’ news and ‘fake’ news that they encounter on Facebook, Twitter, Google or other Internet platforms?

This study seeks to contribute the scholarly debate about self-perception of misinformation, especially regarding people’s assessment of their own ability to spot false information online. Based on a survey on a national, diverse sample of adult Romanians, we found a significant third person effect, especially in the case of distant others. That is, Romanians in our sample believe that fake news affects to a greater degree people in their outer circle than themselves and people in their inner circle. Our research attempts to refine and expand the analysis of people’s self-perception of their capacity to spot misinformation by providing evidence on third person effect and factors that predict it.

The fake news phenomenon: A brief overview of its definition and effects

The recent unprecedented proliferation of fake news, especially of its politically oriented type, has triggered a multitude of reactions, the majority of which consists of concerns over the vulnerability of individuals, organizations (media outlets included) and society, in general, to manipulations and ‘information pollution’ (Wardle and Derakhshan, 2017) that has invaded the public sphere. Multidisciplinary efforts are currently taken worldwide to find solutions to efficiently combat the spread of false content. Among the possible actions that could be used to reduce the proliferation of fake news are the following: promoting media literacy, empowering people (both users and media professionals) to critically evaluate news and other type of information and to spot and report disinformation, supporting fact-checking projects, or urging Internet and social media platforms to refine their software algorithms as to increase the quality of content that they mediate (HLEG, 2018; Lazer et al., 2018). However, these actions to combat fake news seem to assume that most media users are ‘experienced’ and therefore they can recognize fake news when presented to them. Arguably, a large amount of media users, especially social media users (the spread of fake news seems to be intrinsically linked to activity on social media platforms), are ‘naïve’ and, thus, more likely to trust fake content. Furthermore, even more dangerous is people’s perception, especially these naïve media users, that others than themselves are more prone to be affected by the false. Thus, many people are still unaware of the degree to which fake news affects them, because they are overconfident in their potential to disentangle true from false information. Our study aims to test people’s self-perceived ability to detect fake news as well as their perception of other people’s ability to recognize fake news by providing evidence of third person effect and factors that predict it.

As many scholars have argued, ‘fake news’ is a deceptive notion, lacking definitional rigour, which has made it difficult for people to recognize and distinguish between many shapes that false information can take (Bounegru et al., 2017; HLEG, 2018; Wardle and Derakhshan, 2017). Others have suggested that the term is too vague and that it should not be used at all (Tandoc et al., 2018). Some scholars believe that ‘fake news’ is just ‘old wine in new bottles’, and to exacerbate the role of today’s technological environment in which fake news proliferates would be not to be able to see how entrenched fakery (and fake news) has been in the history of journalism (Schudson and Zelizer, 2017; Tandoc et al., 2018). Lazer et al. (2018) describe ‘fake news’ as a type of fabricated content that mimics news media content (p. 1094), whereas Bounegru et al. (2017) consider that online circulation and reception are crucial to fake news and that ‘fake news cannot be fully understood apart from its circulation online’ (p. 8).

Perhaps the most complex and complete definitional framework for ‘fake news’ has been proposed by Claire Wardle (2017) from First Draft.¹ They argue that using the term ‘fake news’ does not do justice to the complexity and diversity of misinformation and disinformation. Therefore, a more accurate and encompassing term to capture these complex phenomena would be information disorders (Wardle and Derakhshan, 2017). To understand what fake news means, Tandoc et al. (2018) critically reviewed 32 academic articles using the term ‘fake news’ and published between 2003 and 2017 and identified six types of fake news: satire, parody, fabrication, manipulation, propaganda and advertising (Tandoc et al., 2018: 5). Maybe, the most relevant definition of this vague concept would be related to whatever people perceive fake news to be. However fluid, self-perception about fake news is what drives people in their digital behaviour.

Our study is consistent with recent research that examined the perceived influence of fake news, from a third person effect perspective. Jang and Kim (2018) showed that people believe that fake news has a greater effect on out-group members than themselves or in-group members, and that partisan identity, social desirability of content and external political efficacy were positive predictors of third person effect. In addition, Ştefăniţă et al. (2018) reported the same phenomenon in Romania, with gender and fake news exposure being the most powerful predictors. This study focuses on a different dimension of the third person effect, namely people’s assessment of self, and others’ ability to detect fake news. This is equally important from the point of view of fighting the fake news phenomenon, in the sense that part of the problem is people’s belief in their ‘gut feeling’ sense in screening the information they consume for the ‘fake’ ingredient.

The third person effect: Components and predictors

Originated in a seminal study of Davison (1983), which suggested that individuals tend to overestimate the influence that media have on the attitudes and behaviours of others and underestimate the same influence on themselves, research on the third person effect forms now a ‘relatively cohesive conceptual structure’ (Golan and Banning, 2008: 208), supported by a large body of empirical studies and developed around two components: the perceptual and the behavioural. The perceptual, or the gaps in perceptions of media effects on self and others, has been systematically studied mainly in negative messages conveyed by television violence (Rojas et al., 1996), pornography (Gunther, 1995), gambling, tobacco, and alcohol advertisements (Banning, 2001; Youn et al., 2000), news stories and coverage (Perloff, 1989; Salwen and Driscoll, 1997) or accusatory political adds (Rucinski and Salmon, 1990).

The behavioural component, that is, the fact that people might be influenced to act based on perceived effects, came more recently under the scholars’ scrutiny. Studies in this particular area provide evidence that media may determine a behavioural intention in a wide array of social life aspects, such as support for restrictions on pornography (Lee and Yang, 1996), television violence (Rojas et al., 1996) or negative political advertising (Salwen and Dupagne, 1999); civil participation, discussion about elections, and support for media censorship (Neuwirth et al., 2002); the likelihood to vote (Golan et al., 2008); behavioural intentions based on exposure to health news (Wei et al., 2008); increased desire to be slim (Park, 2005), and so on.

Beside supporting with empirical evidence the perceptual and the behavioural components, an important direction in the field of the third person effect seeks to understand the complex psychological mechanisms, which determine people to consider that media might have a stronger influence on others than on themselves. Explanations build on motivational as well as on cognitive concepts and theories and concern so far mainly the perceptual component, which is also the focus of our research.

Theories or models such as ego involvement (Perloff, 1989), biased optimism (Brosius and Engel, 1996; Gunther and Mundy, 1993), social comparison theory (Atwood, 1994), attribution theory (Rucinski and Salmon, 1990), or the model of presumed influence (Gunther and Storey, 2003) advance possible patterns for understanding how the perceptual component might occur. The attribution theory, one of the most frequently used within this framework, suggests that an individual ‘tends to attribute his own reactions to the object world, and those of another, when they differ from his own, to personal characteristics in (the other)’ (Heider, 1958: 157). In other words, when it comes to media effects, people tend to adopt a double standard: in what themselves are concerned, they tend to underestimate media effects, since they understand better their own reactions and are aware of the situational factors; in the case of others, when the judgement is based only on perceived characteristics and the contextual factors are largely ignored, media effects are overestimated.

Despite the fact that third person literature is almost overwhelmingly rich, covering many topics and social issues, little is still studied with regard to this effect in the context of the new ‘spectre haunting the world’, fake news. In this context, we propose that

H1: Self perceptions about own ability to detect fake news are higher than perceptions about others’ ability to detect fake news.

H2. Third person effect is stronger when people compare themselves with distant others, than with close others.

According to Rucinski and Salmon (1990), individuals adopt the double standard in order to enhance their ego. The ‘self-serving bias’, which makes us explain our behaviour in terms that put us in good light, could explain why people tend to resist acknowledging media influence on self. In addition, a confirmation bias intervenes when individuals tend to search for information, sources and arguments to confirm their already formed impressions, while opposing views are overlooked. In such cases, the facts are interpreted forcefully to correspond to the beliefs of the person, instead of leading to the updating or changing of these convictions (Bârgăoanu, 2018: 35). When applied to people’s way of relating to the fake news phenomenon, we hypothesize that

H3. People scoring high on the confirmation bias scale are subject to stronger third person effect regarding the ability to detect fake news.

Beyond theoretical models, different factors have been advanced in support of the perceptual component, such as gender and race (David et al., 2002; Lo and Wei, 2002), religiosity (Golan, 2002), education (Peiser and Peter, 2000), the social distance (Jensen and Hurley, 2005), the cultural mores (Lee and Tamborini, 2005) and values (Hong, 2015), and so on. In what concerns education, most studies suggest that better educated people tend to show stronger third person perception (Salwen, 1998; Tiedge et al., 1991), since they tend to consider themselves more knowledgeable, if not generally superior, than others, and to approach media messages more cautiously. In what concerns specifically the fake news, our fourth hypothesis advances that

H4. Higher levels of education are associated with stronger third person effect regarding the ability to detect fake news.

Regularly reading the news and being up-to-date with political affairs might help people resist to persuasive or fabricated media messages, since they feel they can rely on a substantial knowledge which might function as a defence. Politically knowledgeable people are likely to believe that individuals less interested in politics and in news in general might be more permeable to misleading media content. As Rucinski and Salmon (1990) demonstrated, ‘among those with equivalent levels of political interest, higher levels of exposure to newspaper public affairs content increases the third person effect’ (p. 360). In line with these considerations, our fifth hypothesis posits that

H5. The more people are interested in politics, the stronger the third person effect regarding the ability to detect fake news (as they perceive a stronger association between fake news and politics).

Finally, recent studies document the perceptual component of the third person effect in relation with new media use. In what concerns specifically Facebook, researchers provide empirical evidence that individuals consider others as more vulnerable than themselves with respect to the negative effects – such as depression, social isolation, loneliness, decreased self-esteem, uploading of photos that undermine employment opportunities, unwanted contact with dangerous people, and being targeted for cyberbullying and Facebook scams (Kim and Hancock, 2015) – or to the risks (Lev-On, 2016) involved by Facebook use, as well as they believe that low-relevance stories would have a greater impact on others than on themselves (Schweisberger et al., 2014). Moreover, knowledge and perceive knowledge, including in recognizing and fighting fake news, might generate a third person effect (Baek et al., 2013; Jang and Kim, 2018), since new media ‘require greater skills and greater ability to protect oneself against the online risks that are an outcome of the structure of the Internet, the sophistication of the interface, and users’ control over contents’ (Lev-On, 2016: 4).

Method

The design of this research is based on a national survey on a sample of N = 1016 Romanians older than 18 years of age, of which 203 cases were eliminated based on a cleaning procedure using completion time and straightlining criteria. The data were collected by the Survey Sampling International polling company, using a quota sampling on age, education and gender, from 31st July to 10th August 2018. The final sample (N = 813) was diverse from the point of view of gender (52% males), education² (M = 5.97, SD = 1.40), age (M = 40.17, SD = 13.02), political interest (M = 4.26, SD = 1.95),³ and political ideology (M = 6.03, SD = 2.43).⁴

Measures

Third person effect regarding people’s ability to identify disinformation was measured using three items, worded as in Q3 of the Flash Eurobarometer 464: Fake News and Disinformation Online (European Union Data Portal, 2018): ‘How confident or not are you that you / your friends and family / people in general are able to identify news or information that misrepresent reality or is even false?’. Consistent with previous research, we computed two variables, assessing the intensity of the third person effect by subtracting the scores on the ‘close others’ and ‘distant others’ from the ‘self’ variable.

The confirmation bias scale was constructed based on Rassin (2008), using a 14-item scale measured on a 7-point Likert-type scale, of which, consistent with Rassin (2008: 89), a battery of 10 items was kept to form a one single factor scale. CFA using principal component analysis and a varimax rotation confirmed one factor, with loadings from .547 to .732 (α = .838, M = 4.81, SD = .99).

Facebook dependency was measured using a 6-item scale adapted from Ross et al. (2009). The six items loaded on the same factor, with factor loadings ranging from .796 to .971 (α = .908, M = 3.47, SD = 1.71).

Fake news exposure was a self-reported estimation of how often people come across news or information that they believe misrepresent reality or is even false (wording from the Flash Eurobarometer 464 on Fake News and Disinformation Online, European Union Data Portal, 2018), measured on a 5-point scale.

Age, income, education, and political interest were measured with one-item variables, measured continuously (age), on a 10-point scale (income), on an 8-point scale (education) and on a 7-point scale (political interest).

Results

People’s perception about their ability to detect fake news is equally important with their estimation of others’ same ability. This kind of data allows researchers to estimate the degree of vulnerability of people faced with the fake news phenomenon. The third person effect related to people’s ability to detect fake news proved to be significant, when compared to both close and distant others (see Table 1), thus validating H1.

Table 1.

Third person effect perceptions of people’s ability to detect fake news.

	Self	Close others	Distant others
Mean	4.92	4.02	3.07
Standard deviation	1.65	1.60	1.46
T test		T(811) = –16.021	T(812) = –36.198

T tests significant at p < .01.

At the same time, the effect on distant others was significantly more powerful than people’s perceptions about close others (T(812) = –18.58, p < .01) (H2 validated). As previously proceeded (see Salwen and Dupagne, 2001), we run hierarchical regressions to test Hypotheses 3, 4 and 5. In the first step, we introduced the demographics variables (education, income, gender), followed by political interest, fake news exposure, confirmation bias, and Facebook dependency in the second step. In addition, we present results for each variable estimating the ability to detect news (for self, close and distant others).

As far as comparison with close friends and family is concerned, data showed that people’s perception of fake news exposure and political interest were the most powerful predictors of the third person effect about close others, followed by education, Facebook dependency and confirmation bias. When looking at changes in the R², the demographics and personality related variables accounted for almost equal variance in the model. But for estimations of effects on self, the incremental R² value was consistently higher for the individual characteristics block (see Table 2).

Table 2.

Predictors of third person perceptions about people’s ability to detect fake news (for close others), estimated effects on self and on close others.

	Third person perceptions (close others) ß	Estimated effect on self ß	Estimated effect on close others ß
Block 1
Education^a	.120**	.088*	–.056
Income^b	.023	.083	.068
Gender^c	–.053	–.019	.058
Age	–.005	–.028	–.021
R² change	.036	.049	.013
Block 2
Political interest^d	.153**	.268**	.086
Fake news exposure^e	.153**	.127**	–.052
Confirmation bias^f	.105*	.238**	.127**
Facebook dependency^g	–.113**	–.064	.068
R² change	.067	.150	.035
Total R²	.103	.199	.047
Adj. R²	.090	.188	.034
N	569	569	573

The reported ß weights in the hierarchical regressions are final ß weights.

p < .05.

p < .01.

Coded from 1 = low to 8 high.

Coded as 0 = male, 1 = female.

Coded from 1 = low to 7 = high.

Coded from 1 = low to 5 = high.

Coded from 1 = low to 7 = high.

For close others, the analysis validated H3, the third person effect is stronger for higher educated people, as education might enhance feelings of superiority. Confirmation bias was also a significant predictor, the higher people score on the confirmation bias scale (which translates into people being more confident about being able to quickly understand and evaluate any situation), the stronger the third person effect (H4 validated for close others). At the same time, interest in politics significantly increases the intensity of the third person perceptions about close others, as people generally link fake news to political topics (H5 validated for close others).

Even though the wording of people’s estimation of the ability to detect fake news did not refer in any way to social networks, Facebook dependency proved a significant predictor, in the sense that lower Facebook dependency is associated with stronger third person effect. Surprisingly, the direction of this correlation is negative, which could be an effect of what people consider ‘news’ and ‘fake news’ to be. People who are more dependent on Facebook might perceive that ‘news or information’ is content they come across on Facebook, and therefore estimate a distance between them and ‘friends and family’, which are exposed to the same type of content (again, on Facebook). People who are less Facebook dependent might perceive that ‘news’ is rather a more formal content, such as TV news or ‘proper’ online news, but at the same time believe that others are more Facebook dependent and hence more exposed to news through social media, and therefore are more vulnerable to negative effects (stronger third person effect).

Somewhat surprisingly, apart from education, no socio-demographics were significant.

Third person effect by contrast with ‘people in general’ is more consistently explained by the predictors proposed in this study (R² = .143; for comparison R² for close others = .101), following similar patterns of significance: political interest and fake news exposure are the strongest predictors, followed by education, confirmation bias, Facebook dependency and age (age was not significant for close others), thus validating H3, H4 and H5 for distant others (see Table 3).

Table 3.

Predictors of third person perceptions about people’s ability to detect fake news (for distant others), estimated effects on self and on distant others.

	Third person perceptions (distant others) ß	Estimated effect on self ß	Estimated effect on distant others ß
Block 1
Education^a	.158**	.088*	–.125**
Income^b	.066	.083	.009
Gender^c	–.034	–.019	.040
Age	–.098*	–.028	.104*
R² change	.056	.049	.026
Block 2
Political interest^d	.180**	.268**	.045
Fake news exposure^e	.169**	.127**	–.091*
Confirmation bias^f	.130*	.238**	.091*
Facebook dependency^g	–.111**	–.064	.083
R² change	.087	.150	.027
Total R²	.143	.199	.053
Adj. R²	.131	.188	.040
N	569	569	573

The reported ß weights in the hierarchical regressions are final ß weights.

p < .05.

p < .01.

Coded from 1 = low to 8 high.

Coded as 0 = male, 1 = female.

Coded from 1 = low to 7 = high.

Coded from 1 = low to 5 = high.

Coded from 1 = low to 7 = high.

The incremental R² value of the personality characteristics predicting third person effect about people’s ability to detect fake news, when comparing themselves with ‘people, in general’ was greater than the value in the block of demographics.

Comparing third person effect models for close and distant others (Table 4), the only significant difference concerns age, which is a significant predictor in the distant others model, in the sense that younger people are subject to stronger third person effect. Moreover, the second model has a stronger explicative power.

Table 4.

Predictors of third person perceptions about people’s ability to detect fake news for close and distant others.

	Third person perceptions (close others) ß	Third person perceptions (distant others) ß
Block 1
Education^a	.120**	.158**
Income^b	.023	.066
Gender^c	–.053	–.034
Age	–.005	–.098*
R² change	.036	.056
Block 2
Political interest^d	.153**	.180**
Fake news exposure^e	.153**	.169**
Confirmation bias^f	.105*	.130*
Facebook dependency^g	–.113**	–.111**
R² change	.067	.087
Total R²	.103	.143
Adj. R²	.090	.131
N	569	569

The reported ß weights in the hierarchical regressions are final ß weights.

p < .05.

p < .01.

Coded from 1 = low to 8 high.

Coded as 0 = male, 1 = female.

Coded from 1 = low to 7 = high.

Coded from 1 = low to 5 = high.

Coded from 1 = low to 7 = high.

Summing up, people’s ability to detect fake news is subject to third person effects, which are significantly stronger when people compare themselves with distant others than with close others. As predicted, education, political interest and confirmation bias influence these effects in the sense that less Facebook dependent, higher educated people, who are more interested in politics, estimating to encounter misleading information more often are more affected by third person perceptions about the ability to detect fake news. This could have implications for people’s general lack of interest in using fact checkers, when confronted with questionable content, as they feel confident to rely on ‘gut feeling’.

Discussion

That misinformation on social media is damaging societies and erodes public trust in democratic institutions is an already acknowledged phenomenon (European Commission, 2018; Pogue, 2017). According to a recent report (Newman et al., 2018), great percentages of users acquire their news from social media (depending on country), and these news stories undergo the same popularity dynamics as other forms of online contents (such as selfies and photos). In addition, about 54% are concerned about ‘what is real and fake on the Internet’ (Newman et al., 2018: 9).

Our research reveals that the third person effect is part of the mechanism that explains how and why people consume and approach fake news circulating the online media (at least in the context of Romanian information environment). In a nutshell, our data show that the average citizen would use the following line of reasoning when assessing his or her own behaviour in the online environment: ‘I am aware that there are many fake news around, but it is surely them – my close friends and people in my network – who are mainly affected, as I am generally more aware’. Studying the third person effect is particularly important, because it helps researchers understand how some key variables – such as education, age, political interest and so on – might influence people’s opinions and attitudes on important topics.

Third person effects related to fake news consumption has been previously investigated, but researchers have so far reported only on people’s perception about being influenced by fake news (Jang and Kim, 2018; Ştefăniţă et al., 2018). Thus, people’s assessment of their own capacity to detect fake news was still subject to investigation. This study shows that, at least in Romania, there is a significant third person effect related to people’s ability to detect fake news; thus, people believe close others to be more influenced than themselves, and less influenced than distant others. This is one important finding, in the context of the structural struggle to address the ‘fake news’ phenomenon at the (European) policy level, showing a lower relevance of fact checkers than previously hoped: people might largely rely on gut feeling to distinguish facts from misleading information, as they mostly believe only others to be victims of disinformation.

As far as the predictors of such an effect are concerned, among socio-demographics education proved to be significant: educated people tend to perceive the gap between themselves and others greater than low educated people. Age only matters when positioning one’s self in comparison with distant others: young people tend to perceive a greater gap than older people.

Moreover, data showed that non socio-demographic parameters are more influential in the third person effect, and explain more (greater R² change) the predictive model for both close and distant others. Among them, Facebook dependency negatively correlates with third person perceptions, which is at first counter-intuitive. We expected Facebook dependency to be positively correlated with the intensity of the third person effect, as in Romania, Facebook is the most used social network site (Bârgăoanu and Radu, 2018), including news consumption. The negative correlation suggests an alternative explanation: people who are less dependent on Facebook tend to estimate a wider gap between themselves and (both close and distant) others, as they might believe others to be more Facebook dependent and thus more exposed to news through social networks, as opposed to themselves who rely more on classic/traditional (and thus presumably more reliable) media outlets for news consumption. It is still subject to further investigation whether or not people associate ‘fake news’ with social networks or with a broader range of media outlets. Even though third person effect has been measured about ‘news or information that misrepresent reality or is even false’ in general, results suggest that most of the people might intuitively associate ‘news and information’ with social media, especially linking ‘fake news’ with online content (this supposition should be further tested). At the same time, concerns are still raised with regard to what people consider ‘misleading information’, ‘false information’ or ‘fake news’ to be.

Political interest is an important predictor of the third person effect regarding fake news detection, as there is a common belief that fake news is mostly about political matters. This claim is supported by the background of the ‘fake news’ concept, often perceived as a new phenomenon that emerged with the Trump winning elections in the United States and the Brexit vote in the United Kingdom. Our study reveals that political interest is in fact the most powerful predictor for the third person effect (for both close and distant others models), in the sense that the more interested people are in politics, the wider the gap they perceive between themselves and others in terms of detecting misleading information.

Confirmation bias and fake news exposure are two other significant predictors which, as expected, influence the intensity of the third person effect: this kind of effect is nested in people’s reliable first impression (confirmation bias) of fake news (exposure). The higher people score on the confirmation bias scale, the greater the effect, the higher the perceived fake news exposure, the greater the effect.

The results of this study should be interpreted, keeping in mind some limitations. Even though the sample was not a representative one, it was heterogeneous enough to allow testing for various predictors and to suggest possible interpretations of how people estimate their own ability and others’ to detect misleading and even fake information. However, generalization of results should be taken cautiously, also due to the socio-political context of Romania at the time data were collected (i.e. internal political turmoil, including street protests on political issues). Even though the data did not take into account context related variables, the issue of fake news is mostly primed during periods of turmoil, as people pay closer attention to media news on the one hand, and are more likely to come across vivid discussions about the issue in online media debates. In addition, even though the questions measuring people’s perceptions about own ability to detect misleading content was not worded specifically to address online (social media) disinformation, we believe that people’s representations of the so-called news phenomenon are intrinsically linked to such content, especially since Facebook dependency proved to be a significant predictor.

Despite efforts done by Facebook and other social media platforms to fight disinformation, the online environment is still full of misleading stories. Both communication scholars and journalists report that fact-checking is limited or is simply not working (see Levin, 2017), that the fake news phenomenon is becoming unstoppable (Ghosh and Scott, 2018), and that ‘the absolute level of interaction with misinformation remains high’ (Allcott et al., 2019: 3). Studies focused on the third person effect associated with fake news are worth using to assess and design some viable means to tackle this phenomenon. Our research reveals, among other aspects, that fact-checking agencies might not be the right answer to this kind of problem, especially due to people’s internal mechanisms of assessing both detection and effects of fake news on themselves and on others. Third person effect mechanisms work in various ways: as people place themselves on the higher half of the scale assessing self-perception of the ability to detect fake content based on gut feeling, they might see no point in checking the accuracy of information they consume.

Conclusion

In a world saturated with proliferating, multiple truths and characterized by an instant consumption of information and emotions, it is increasingly difficult to question the validity of the information or the credibility of its source. People often tend to rely on what they already believe, know and feel, as this comforts their convictions and their perspective on reality. This might be particularly problematic in the context of their social networks exposure. People often stop questioning the ‘truth’ of the content they consume online, as they tend to get engaged in interminable discussion about one side or another of a story, whichever in fact confirms their own stereotypes, biases, beliefs, or values. The truthfulness of a matter becomes secondary, as long as the matter itself closely resonates or prime prior attitudes in people’s minds.

As a matter of fact, from casting doubt or concern on a specific – often political – aspect, to initiate conspiracy theories, ‘fake news’ might be particularly harmful to our society. They might polarize the society, intensify or suppress social conflicts, or distract people’s attention from important issues. In this context, people’s self-perceived competency to detect fake news, empirically tested by our study, and the strong third person effect found with regard to the same ability when it comes to closed and distant others might be problematic. Individuals risk to have a distorted perception of reality, be it social or political, or to be more easily manipulated. More specifically, in the case, for example, of the political interest, which proved here to be the most powerful predictor for the third person effect (for both close and distant others models), this might question the very foundations of the democratic process, since people’s trust and engagement into politics might be jeopardized by false stories. By emphasizing the effects that ‘fake news’ might have on individuals, our article contributes to a better understanding of how to address them in the near future. Even though third person effect regarding fake news detection might be culturally bound, we believe that studies like this are necessary steps towards understanding the more intimate mechanisms through which people are affected by fake news. We share concerns with both the academic and professional communities interested in finding viable solutions to the proliferation of fake news in the new media environment, and feel the need to emphasize that efficiently addressing a social issue is not possible without understanding the mechanisms and effects of the phenomenon under scrutiny. We plea for comparative research that should address this complex phenomenon, as one media and political context (in this case the Romanian one) might not offer enough general results to suggest policy making recommendations.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article:The author(s) received financial support for this research from the Doctoral School in Communication, National University of Political Studies and Public Administration, Romania.

ORCID iD

Nicoleta Corbu

Notes

References

Allcott

Matthew

(2019) Trends in the diffusion of misinformation on social media. Working Paper 25500, January. National Bureau of Economic Research. Available at: https://www.nber.org/papers/w25500.pdf (accessed 4 April 2019).

Atwood

(1994) Illusions of media power: The third person effect’. Journalism & Mass Communication Quarterly 71: 269–281.

Baek

Kim

E-M

Bae

(2013) My privacy is okay, but theirs is endangered: Why comparative optimism matters in online privacy concerns. Computers in Human Behavior 31: 48–56.

Ball

(2017) Post-Truth: How Bullshit Conquered the World. London: Biteback Publishing.

Banning

(2001) Do you see what I see? Third person effects on public communication through self-esteem, social stigma, and product use. Mass Communication & Society 4(2): 127–147.

Bârgăoanu

(2018) Fakenews: Noua cursă a înarmării (#Fakenews: The new arms race). Bucharest: Evrika Publishing.

Bârgăoanu

Radu

(2018) Fake news or disinformation 2.0? Some insights into Romanians’ digital behaviour. Romanian Journal of European Affairs 18(1): 24–38.

Bounegru

Gray

Venturini

, et al. (2017) A field guide to fake news and other information disorders. Amsterdam: Public Data Lab. Available at: https://fakenews.publicdatalab.org/ (accessed 10 October 2018).

Brosius

H-B

Engel

(1996) The causes of third person effects: Unrealistic optimism, impersonal impact, or generalized negative attitudes towards media influence? International Journal of Public Opinion Research 8: 142–162.

10.

David

Morrison

Johnson

, et al. (2002) Body image, race and fashion models: Social distance and social identification in third person effects. Communication Research 29(2): 270–294.

11.

Davison

(1983) The third-person effect in communication. Public Opinion Quarterly 47: 1–15.

12.

European Commission (2018) A multi-dimensional approach to disinformation: Final report of the high-level expert group on fake news and online disinformation. Available at: https://op.europa.eu/en/publication-detail/-/publication/6ef4df8b-4cea-11e8-be1d-01aa75ed71a1

13.

European Union Data Portal (2018) Flash Eurobarometer 464: Fake news and disinformation online. Available at: https://data.europa.eu/euodp/data/dataset/S2183_464_ENG.

14.

Ghosh

Scott

(2018) Digital deceit: The technologies behind precision propaganda on the internet. Available at: https://www.newamerica.org/public-interest-technology/policy-papers/digitaldeceit/ (accessed 1 April 2019).

15.

Golan

(2002) Religiosity and the third person effect. Journal of Media and Religion 1(2): 105–120.

16.

Golan

Banning

(2008) Exploring a link between the third person effect and the theory of reasoned action: Beneficial ads and social expectations. American Behavioral Scientist 52(2): 208–224.

17.

Golan

Banning

Lisa

(2008) Likelihood to vote, candidate choice, and the third person effect behavioral implications of political advertising in the 2004 presidential election. American Behavioral Scientist 52(2): 278–290.

18.

Gunther

(1995) Overrating the X-rating: The third person perception and support of censorship of pornography. Journal of Communication 45(1): 27–38.

19.

Gunther

Mundy

(1993) Biased optimism and the third person effect. Journalism Quarterly 70: 58–67.

20.

Gunther

Storey

(2003) The influence of presumed influence. Journal of Communication 53(2): 199–215.

21.

Heider

(1958) The Psychology of Interpersonal Relations. New York: John Wiley & Sons.

22.

HLEG (2018) A multi-dimensional approach to disinformation: Report of the independent High Level Expert Group (HLEG) on fake news and online disinformation. Available at: http://ec.europa.eu/newsroom/dae/document.cfm?doc_id=50271 (accessed 10 October 2018).

23.

Hong

(2015) Do cultural values matter? A cross-cultural study of the third person effect and support for the regulation of violent video games. Journal of Cross-Cultural Psychology 46(7): 964–976.

24.

Jang

Kim

(2018) Third person effects of fake news: Fake news regulation and media literacy interventions. Computers in Human Behavior 80: 295–302.

25.

Jensen

Hurley

(2005) Third person effects and the environment: Social distance, social desirability, and presumed behavior. Journal of Communication 55: 242–256.

26.

Kim

Hancock

(2015) Optimistic bias and Facebook use: Self-other discrepancies about potential risks and benefits of Facebook use. Cyberpsychology, Behavior and Social Networking 18(4): 214–220.

27.

Lazer

DMJ

Baum

Benkler

, et al. (2018) The science of fake news: Addressing fake news requires a multidisciplinary effort. Science 359: 1094–1096.

28.

Lee

Tamborini

(2005) Third-person effect and internet pornography: The influence of collectivism and internet self-efficacy. Journal of Communication 55: 292–310.

29.

Lee

Yang

(1996) Third person perception and support for censorship of sexually explicit visual content: A Korean case. Paper presented at the association for education in journalism and mass communication, Anaheim, CA, August.

30.

Levin

(2017) Facebook promised to tackle fake news: But the evidence shows it’s not working. The Guardian, 16 May. https://www.theguardian.com/technology/2017/may/16/facebook-fake-news-tools-not-working (accessed 1 April 2019).

31.

Lev-On

(2016) The third person effect on Facebook: The significance of perceived proficiency. Telematics and Informatics 34(4): 252–260.

32.

V-H

Wei

(2002) TPE, gender and pornography on the Internet. Journal of Broadcasting & Electronic Media 46: 13–33.

33.

Neuwirth

Frederick

Mayo

(2002) Person-effects and heuristic-systematic processing. Communication Research 29(3): 320–359.

34.

Newman

Fletcher

Kalogeropoulos

, et al. (2018) Reuters Institute Digital News Report 2018. Reuters Institute. Available at: https://reutersinstitute.politics.ox.ac.uk/sites/default/files/digital-news-report-2018.pdf.

35.

Park

S-Y

(2005) The influence of presumed media influence on women’s desire to be thin. Communication Research 32: 594–614.

36.

Peiser

Peter

(2000) Third person perception of television-viewing behavior. Journal of Communication 50(1): 25–45.

37.

Perloff

(1989) Ego-involvement and the TPE of televised news coverage. Communication Research 16: 236–262.

38.

Pogue

(2017) How to stamp out fake news. Scientific American 316(2): 24–24.

39.

Rassin

(2008) Individual differences in the susceptibility to confirmation bias. The Netherlands Journal of Psychology 64(2): 87–93.

40.

Rojas

Shah

Faber

(1996) For the good of others: Censorship and the third person effect. International Journal of Public Opinion Research 8: 163–186.

41.

Ross

Orr

Sisic

, et al. (2009) Personality and motivations associated with Facebook use. Computers in Human Behavior 25(2): 578–586.

42.

Rucinski

Salmon

(1990) The ‘other’ as vulnerable voter: A study of the third-person effect in the 1988 U.S. Presidential campaign. International Journal of Public Opinion Research 2: 343–368.

43.

Salwen

(1998) Perceptions of media influence and support for censorship: The third person effect in the 1996 presidential election. Communication Research 25: 259–285.

44.

Salwen

Driscoll

(1997) Consequences of third person perception in support of press restrictions in the O. J. Simpson trial. Journal of Communication 47(2): 60–78.

45.

Salwen

Dupagne

(1999) The third person effect: Perceptions of the media’s influence and immoral consequences. Communication Research 26(5): 523–550.

46.

Salwen

Dupagne

(2001) Third person perception of television violence: The role of self-perceived knowledge. Media Psychology 3(3): 211–236.

47.

Schudson

Zelizer

(2017) Fake news in context. In: Understanding and Addressing the Disinformation Ecosystem. Annenberg School for Communication, pp. 1–4. Available at: https://firstdraftnews.org/wp-content/uploads/2018/03/The-Disinformation-Ecosystem-20180207-v4.pdf (accessed 10 October 2018).

48.

Schweisberger

Billinson

Makana Chock

(2014) Facebook, the third person effect, and the differential impact hypothesis. Journal of Computed-Mediated Communication 19(3): 403–413.

49.

Ştefăniţă

Corbu

Buturoiu

(2018) Fake news and the third-person effect: They are more influenced than me and you. Journal of Media Research 11(3): 5–23.

50.

Tandoc

Jr Lim

Ling

(2018) Defining ‘fake news’: A typology of scholarly definitions. Digital Journalism 6(20): 137–153.

51.

Tiedge

Silverblatt

Havice

, et al. (1991) Discrepancy between perceived first-person and perceived third-person mass media effects. Journalism Quarterly 68: 141–154.

52.

Wardle

(2017) Fake news: It’s complicated. First Draft, 16 February. Available at: https://firstdraftnews.org/fake-news-complicated/ (accessed 10 October 2018).

53.

Wardle

Derakhshan

(2017) Information disorder: Toward an interdisciplinary framework for research and policy making. Council of Europe Report. Available at: https://rm.coe.int/information-disorder-report-november-2017/1680764666.

54.

Wei

V-H

H-Y

(2008) Third-person effects of health news exploring the relationships among media exposure, presumed media influence, and behavioral intentions. American Behavioral Scientist 52(2): 261–277.

55.

Youn

Faber

Shah

(2000) Restricting gambling advertising and the third-person effect. Psychology and Marketing 17(7): 633–649.