See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/345737746

Third‐Person Effect

Article · September 2020

DOI: 10.1002/9781119011071.iemp0130

CITATIONS

0 READS

220

1 author:

Mina Tsay-Vogel

Boston University

31 PUBLICATIONS   750 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Mina Tsay-Vogel on 09 March 2021.

The user has requested enhancement of the downloaded file.

Third-Person Effect MINA TSAY-VOGEL Boston University, USA

Third-person effect (TPE) is the tendency for people to perceive others as being consid- erably more influenced by mass media messages as compared to themselves (Davison, 1983). In other words, an individual believes that messages do not have their great- est effect “on me or you, but on them—the third persons” (Davison, 1983, p. 3). This perceptual bias to exaggerate the impact of mass communication on public opinion can have consequential effects on actions, often referred to as the behavioral compo- nent of TPE. An assumption made by TPE is that estimates of media influence on self and others are separate entities such that people have the potential to discriminate the impact of mass media on others and that on themselves (Perloff, 1999). Scholars have also referred to this seeming discrepancy in media influence as a perceptual distortion, implying that most individuals are willing to acknowledge this logical inconsistency (Tiedge, Silverblatt, Havice, & Rosenfeld, 1991).

Applications of TPE

The concept of TPE has been applied to a multitude of media channels, message types, media genres, and media sources. Early support for TPE (Davison, 1983) revealed that when asked to assess the influence of persuasive communication, people were more likely to report that everyone else is more easily influenced by media messages than themselves when evaluating the effect of television advertising on children and the impact of campaign themes, early presidential primaries, and campaign advertising on voters. The judgment of greater media effects on others than the self has also received empirical support in various contexts including pornography, defamatory news, news about controversial political issues, television violence, product commercials, and rap music (Perloff, 1993, 1999). Meta-analyses of TPE have yielded average effect sizes (r) of .50 (Paul, Salwen, & Dupagne, 2000) and .31 (Sun, Pan, & Shen, 2008) for the perceptual component of the hypothesis.

Recent work has applied TPE to new media to understand self–other disparities in the estimation of media impact in the context of social media (Golan & Lim, 2016; Schweisberger, Billinson, & Chock, 2014; Tsay-Vogel, 2016). Specifically, users reported that Facebook exerts stronger effects on others than themselves, yielding patterns consistent with the TPE hypothesis (Tsay-Vogel, 2016). Self–other differences in perceived media vulnerability have also been investigated in numerous online contexts as they pertain to the presumed influence of privacy, brand marketing,

The International Encyclopedia of Media Psychology. Jan Van den Bulck (Editor-in-Chief), David Ewoldsen, Marie-Louise Mares, and Erica Scharrer (Associate Editors). © 2020 John Wiley & Sons, Inc. Published 2020 by John Wiley & Sons, Inc. DOI: 10.1002/9781119011071.iemp0130

2 THIRD- PERSON EFFECT

social media posts, recruitment for political and social causes, blogs, news, and health information.

Theoretical mechanisms of TPE

The theoretical underpinnings of TPE have commonly focused on why social compar- isons and contrasts are made. According to attribution theory (Heider, 1958), individ- uals strive to understand relevant events and behaviors by making judgments based on observations. However, there are instances which result in discrepancies between attributes to self and others such that people tend to ascribe their own actions to sit- uational contexts, whereas observers are more likely to ascribe the same actions to personal characteristics. This actor–observer difference in causal attributions of behav- ior is referred to as attributional error and helps to explain why individuals believe they are more informed about and resistant to the persuasive appeal (e.g., source intention) of a media message than others as they are inclined to attribute dispositional shortcom- ings to other people (e.g., credulity or lack of knowledge) (Gunther, 1991).

Third-person effect is fundamentally grounded on the motivation to preserve a posi- tive self-image based on social desirability, also referred to as the self-serving bias (Gun- ther & Mundy, 1993). People tend to engage in downward comparisons (i.e., comparing themselves to those worse off) in order to maintain or enhance self-esteem (Gunther, 1991). Individuals may also make social comparisons to the point of upholding unre- alistic positive images of themselves such that they believe they are less susceptible to negative outcomes and conversely, more likely to experience positive ones as compared to others (Gunther & Mundy, 1993). This biased optimism can further explain why people deem themselves as being more impervious to the negative influence of persua- sive messages as they believe they are wiser than average people. However, this effect is assumed to only occur under conditions in which the perception of oneself as being less influenced by media has self-serving benefits or is ego-enhancing.

In addition to self-enhancement motivation as a psychological driver of TPE, Perloff (1993) suggests that media effects schemas, or mental structures of preconceived ideas about the media and the audience, partially explain the differential estimated effects of media on self versus others. Based on traditional conceptions of mass communication informed by the hypodermic needle model, individuals have a tendency to believe that media are powerful agents of change and that audience members are normally passive, unaware, and susceptible to media influence. In turn, people would logically judge others as falling prey to media persuasion as compared to themselves.

Scholars also suggest that perceived media exposure contributes to the assessment of media impact (Golan & Lim, 2016; McLeod, Eveland, & Nathanson, 1997; Tsay-Vogel, 2016). It is possible that people use simple heuristics of media exposure to guide their evaluation of media effects. For example, individuals may believe that if others are consuming media, they will inevitably be influenced by such messages, consistent with early theories of mass communication as having direct and prevailing effects on audiences.

THIRD- PERSON EFFECT 3

Conditions affecting TPE

Research has shown that a number of factors influence the degree to which TPE is facilitated or minimized. Perceived message desirability has been found to be negatively associated with the likelihood and magnitude of TPE (Duck & Mullin, 1995; Gunther & Mundy, 1993). Specifically, individuals have greater propensity to report that others are more affected by media messages than they are themselves when judging the impact of content that has harmful consequences (e.g., advertisements promoting diet pills, news stories in support of extreme right-wing political parties, and messages containing themes of violence, sexism, and racism). More pronounced TPE for antisocial messages or those not consonant with social norms are explained in light of the self-serving inclination to see oneself in a positive light (e.g., being intelligent enough to resist the effects of negative messages) and judge vulnerable others as being more likely to be duped. Furthermore, media formats that are generally perceived as “undesirable to be influenced by,” such as commercial advertisements, result in greater TPE than those lacking such negative connotation, such as news and public service announcements (Gunther & Mundy, 1993). In essence, denying the effect of media especially in the case when such messages or formats are deemed personally disadvantageous helps to project a superior self-image.

Conversely, in the case of socially desirable messages or those that advocate personally beneficial outcomes, scholars suggest that people are less likely to exhibit TPE and, instead, acknowledge stronger media influence on themselves than others. This phenomenon is referred to as first-person effect (FPE) or reversed TPE (Tiedge et al., 1991). FPE is also consistent with the self-enhancement perspective as the impact of prosocial messages is considered positive, highly regarded, and ego-boosting (Duck & Mullin, 1995; Gunther & Mundy, 1993). While there is robust support for antisocial messages producing TPE, studies examining FPE have yielded mixed results. There is some evidence corroborating that people who deem public service campaigns as desirable report greater self-influence (Duck & Mullin, 1995), however, most studies examining discrepancies in perceived media effects between self and others in the context of messages that convey positive social values (e.g., wearing seatbelts, adhering to traffic safety, showing resistance to antisocial temptations, and displaying concern for others) have revealed only diminished TPE and no FPE (Gunther & Mundy, 1993). The lack of support for FPE may be explained by the assumed perception of social desirability of media messages as opinions about issues or message desirability are not explicitly measured (Paul et al., 2000). It is possible that in these studies, desirable messages were not perceived as personally beneficial as they were presumed to be. Furthermore, while messages associated with violence, sexism, or racism are generally considered antisocial in nature, other message types such as news or advertising are more ambiguous in terms of message desirability. Individuals may also be inherently hesitant to admit to any personal influence as doing so would be ego-threatening. Therefore, even if people comply with positive messages, the mere acknowledgment of being affected by media is perhaps undesirable (Gunther & Mundy, 1993), potentially accounting for the less readily documented FPE.

4 THIRD- PERSON EFFECT

Another factor that can affect the magnitude of TPE pertains to the source of the message. Specifically, if the message communicator is judged to be negatively biased or if the audience believes there is persuasive intent behind the message, the disparity in perceptions of media effects between self and others increases (Cohen, Mutz, Price, & Gunther, 1988; Gunther, 1991; Gunther & Mundy, 1993). The influence of source bias on TPE has been examined in the context of libelous, defamatory, and sensationalistic news. Specifically, research has found that the self–other discrepancy in perceived influence of media was greater when the story was delivered by a biased source such as The National Enquirer, as compared to an unbiased source such as The New York Times (Gunther, 1991).

In addition to message desirability and source bias, the strength of TPE also hinges on the perceived social distance between the self and comparison other (Perloff, 1999). Social distance takes into consideration one’s perceived similarity, familiarity, and identification with the referent group (Perloff, 1993). According to the social distance corollary, as the perceived difference between self and others increases (i.e., referent others are more psychologically removed from the self), TPE is more pronounced. Explanations of this phenomenon have been informed by the process of downward social comparison and the evaluation of distant others as an outgroup. Additionally, if target referents have attributes that align with the prototype of being weak and vul- nerable to media influence, a much larger self–other difference in presumed message effects results (Sun et al., 2008). Considerable support for the social distance corollary has been documented, showing that people are more likely to estimate stronger effects of media on others when evaluating generalized and hypothetical third-persons (e.g., the public at large or people defined in broad, vague, and global terms) versus close others (e.g., friends) (Gunther, 1991).

Other individual difference factors that influence the strength of TPE include ego-involvement, knowledge, and age. When individuals are more ego-involved in a message (i.e., find the topic to be personally important or relevant to their self-concept), the tendency to perceive media as affecting others more than themselves becomes magnified (Perloff, 1999, 2003). Specifically, highly ego-involved individuals are apt to view media sources as negatively biased due to having strong and extreme attitudes about an issue. This fosters not only an increased disparity in perceptions of media effects between self and others, but also the hostile media phenomenon, the belief that neutral messages are biased against one’s own perspective (Vallone, Ross, & Lepper, 1985). Recent work in the context of social media reveals that people tend to evaluate personally relevant news on Facebook as exerting stronger influence on themselves as compared to non-personally relevant news (Schweisberger et al., 2014). Moreover, subjects in the study judged stories of low relevance to have greater effects on others than themselves, supporting the general patterns of TPE. TPE is also likely to occur among individuals with better education and those older in age (Tiedge et al., 1991). People with greater knowledge perhaps believe they are more impervious to mass media’s influence as they see themselves as having the intelligence and expertise to resist such effects, compared to others who lack this proficiency. Older people may assess mass communication as exerting minimal effects on themselves in light of their

THIRD- PERSON EFFECT 5

confidence in countering such influence due to life experiences and accessibility of social attitudes.

Criticisms of TPE

Although there has been robust empirical support for TPE, scholars have questioned whether the disparity in perceived media influence between self and others is an outcome of a measurement artifact (Perloff, 1999). Researchers have typically used two methods to empirically test self–other discrepancies in the evaluation of media influence. In particular, experiments have exposed individuals to certain types of media messages and then asked subjects to evaluate the influence of such content on their own attitudes and the attitudes of others. Alternatively, participants have been surveyed to simply report their assessments of the impact of particular types of messages on themselves and those on others without direct media exposure. One concern is that TPE emerges because of the order in which questions about communication effects on others versus self are presented. Specifically, TPE may be a result of a primacy effect in which people are asked to evaluate the influence of media on others before judging that on themselves. This phenomenon has been discounted by studies counterbalancing the order of self–other questions and still revealing findings that support TPE (Tiedge et al., 1991). Additional concerns over TPE as a result of a perceptual contrast (i.e., natural inclination for responses to favor a disparity in self–other effects due to the presentation of back-to-back questions about the effects of media on self vs. others) have also been disconfirmed (Price & Tewksbury, 1996).

Scholars have also questioned the degree to which TPE stems from a psychological distortion (Perloff, 1993, 1999). This debate raises concern about whether people are overestimating the effects of media on others or underestimating media influ- ence on themselves. Overestimation can be explained by schemas associated with the persuasive influence of media on gullible and susceptible audiences, whereas underestimation can be linked to one’s inability to understand their psychological functioning or motivation to boost self-worth via an illusion of invulnerability (Perloff, 1993). Mixed results of the nature of this psychological distortion have emerged when comparing perceptions of media effects with actual opinion change or opinions of equivalent groups. Whereas some evidence supports the notion that people accurately report the effects of mass media on themselves while overestimating their impact on others (Gunther, 1991), scholars have also found support for the underestimation of self-influence (Cohen et al., 1988). Despite these inconsistent patterns, the existence and nature of the distortion of self–other disparities in the estimation of media effects still warrant greater investigation.

Consequences of TPE

According to the behavioral component of the TPE hypothesis, the bias in the assess- ment of stronger media influence on others rather than the self should likely induce

6 THIRD- PERSON EFFECT

actions, such as support for censorship of media content or change in public policies (Davison, 1983). Although substantial research has focused on the perceptual aspect of TPE, the consequences that result from self–other discrepancies in perceived media effects have received less attention (Paul et al., 2000; Sun et al., 2008). Among the studies that have examined implications of TPE, most yield support for TPE in propelling the restriction of media messages that are considered to produce harmful outcomes (e.g., pornography, violence, and sexual content on television, rap music, and advertisements promoting cigarettes, alcohol, and gambling, with mixed results for news and political content [Perloff, 1999]). Recently, scholars have found support for corrective actions as behavioral outcomes of TPE in the context of social media (Golan & Lim, 2016). Specifically, the perception of others as being vulnerable to the influence of ISIS (Islamic State in Iraq and Syria) online recruitment was positively associated with people’s like- lihood to participate in social media activism or corrective behaviors, such as sharing anti-ISIS content on their social media pages.

In addition to supporting the link between third-person perception and restrictive or corrective actions, research has also examined the effects of TPE on people’s willingness to publicly express their opinions and perceptions of public opinion (Perloff, 1999). The latter implication is known as the persuasive press inference, which suggests that individuals make evaluations about public opinion based on their judgment of news coverage and the assumptions underlying the persuasive impact of that coverage on others (Gunther, 1998).

While much of TPE research is survey based, concerns over the causal order of TPE impacting behavioral outcomes have been raised (Perloff, 1999). It is reasonable that one’s motivation to support government censorship of media materials due to strong apprehensions about the influence of such messages on other people causes TPE, instead of the reverse. Furthermore, in studies that have examined behavioral outcomes of TPE, actual behaviors were not measured, but rather the tendency to exhibit these behaviors was assessed.

Due to theoretical and methodological limitations surrounding the behavioral com- ponent of TPE and the scarcity of behavioral investigations (Perloff, 1999), it has not been feasible for researchers to conduct a meta-analysis to capture the average effect size of the behavioral outcomes of third-person perception (Paul et al., 2000; Sun et al., 2008). Theoretical advances and more comprehensive empirical examination of the behavioral implications of TPE are necessary before a meta-analytical review of these outcomes is possible.

SEE ALSO: Hostile Media Effect; Influence of Presumed Media Influence; Question Wording and Item Formulation; Social Comparison Theory; Spiral of Silence; Survey Methods, Traditional, and Public Opinion Polling

References

Cohen, J., Mutz, D., Price, V., & Gunther, A. C. (1988). Perceived impact on defamation: An experiment on third-person effects. Pub