Reading I NEED HELP ON MY HOMEWORK 3

Chapter9thirdpersoneffect.pdf

Home >Reading homework help >Reading I NEED HELP ON MY HOMEWORK 3

c h a p t e r n I n e

Third-Person Theory

The Third-Person Effect (TPE) refers to a contrast in beliefs where individuals believe that other people (third persons) are more vulnerable to media influence than they are (first person). Thus the foundation of TPE is a “self-serving” percep- tion whereby people think the media exert powerful influences but only on other people, not on themselves. This allows people to complain about the media and call for regulation of harmful content so as to control the media exposures of other people. At the same time, it excuses themselves from having to take responsibility for possible negative consequences of their own exposures as long as people tell themselves that the media have little or no influence on them personally.

When this idea was introduced into the scholarly literature in the early 1980s, it was generally regarded as an intriguing effect but not a theory. Over time as the idea attracted media effects scholars who developed explanations for why this effect continually occurs, it grew into a system of explanation about media influ- ence that now includes work on third-person perceptions (TPPs), first-person perceptions (FPPs), and has even expanded to be called by some Presumed Media Influence (Tal-Or, Tsfati, & Gunther, 2009).

202 | analysis of core theories

Original Conceptualization of the Theory

Conceptual Background

Shortly after World War II, sociologist W. Phillips Davison began studying how propaganda was used during the war to influence public opinion. In one study he examined an incident where the Japanese tried to discourage black American soldiers from fighting at Iwo Jima by dropping leaflets on their encampments. The leaflets tried to convince the black soldiers that they should surrender or desert. Although there was no indication that the leaflets had any effect on the soldiers, the incident preceded a substantial reshuffle among the officers and the unit was withdrawn the next day. Davison was puzzled that the commanders believed the soldiers were influenced by the leaflets even though the black soldiers were not (Davison, 1983).

Several years later while interviewing West German journalists to determine the influence of the press on foreign policy, Davison asked the journalists to esti- mate the influence their editorials had on readers. He found that the journalists typically said, “The editorials have little effect on people like you and me, but the ordinary reader is likely to be influenced quite a lot” (Davison, 1983). However, Davison could find no evidence to support these claims made by those journalists. Davison started going through the public opinion literature to see if he could find other instances of this pattern where people exhibited beliefs that other people (third persons) were influenced by media messages but they (first person) were not. Then in the early 1980s, Davison conducted a series of four surveys to test this pattern. In each survey he asked between 25 and 35 participants to estimate the influence of persuasive communication on themselves and others. Participants estimated self-other effects for (1) a campaign theme on gubernatorial vote choice, (2) television advertising on children, (3) the results of early presidential prima- ries on vote choice, and (4) campaign messages on presidential vote choice. On average they estimated (1) other New York voters were more influenced by cam- paign themes than they were personally, (2) other children were more influenced by television advertising than they had been personally when they were children, (3) others were more influenced by the results of early presidential primaries than they were personally, and (4) others were more influenced by campaign advertise- ments than they were personally. The findings from this series of informal surveys all showed support for the same pattern where respondents consistently believed that other people were more strongly affected by the media than they were person- ally (Davison, 1983).

third-person theory | 203

Introduction of the Theory

Davison introduced the idea of TPE into the scholarly literature in an article enti- tled “The third-person effect in communication,” which was published in Public Opinion Quarterly in 1983. In this article, he presented the results from his four surveys; then in a review of published studies that tested beliefs about media influ- ence, he reinterpreted their findings to show they were in line with a TPE. He also argued that TPE was related to but distinct from other effects such as mispercep- tion of public opinion, pluralistic ignorance, and spiral of silence.

It is important to note that Davison did not refer to TPE as a theory. Instead Davison referred to this pattern as “the third-person effect” and “the third-person hypothesis,” which were the names typically used by subsequent scholars (Gunther, 1995; Hoornes & Ruiter, 1996; Huh, Delorme, & Reid, 2004; McLeod, Eveland, & Nathanson, 1997; Price, Tewksbury, & Huang, 1998; Shah, Faber, & Youn, 1999; Sun, Pan, & Shen, 2008; Tiedge, Silverblatt, Havice, & Rosenfeld, 1991). However, it has also been referred to as “third-person con- ceptualization” (Cohen & Davis, 1991), “third-person effect model” (Gunther, 1995), and “third-person phenomenon” (Gunther & Thorson, 1992). Some scholars are careful not to refer to it as a theory. For example, in the first major review of TPE, Perloff (1999) said, “It is important to emphasize that the TPE is not a theory of public opinion, but rather a hypothesis or series of assertions about perceptions of public opinion and their effects” (p. 355) but he did not explain why he made this assertion.

While Davison’s introduction of TPE in 1983 exhibits some characteristics of a theory (definitions of key concepts, a general proposition that can be tested through operationalization, and grounding in scholarly literatures), it is missing others (additional propositions that could form a set that provides a systematic explanation of some phenomenon). But this was enough to attract many scholars to this idea of TPE, and over the next few decades those scholars have elaborated many of Davison’s ideas and published tests of TPE in scholarly journals. This lit- erature now presents a system of explanation with multiple propositions that can be tested, and when they are tested have been found to be supported.

In this book, TPE is regarded as a theory because it has evolved to meet the criteria for theory as laid out in Chapter 2. It is a system of explanation that features a set of propositions and constructs; the constructs are measurable; the claims made in those propositions are testable and falsifiable by operationalizing them into hypotheses as evidenced by an already large and growing literature of empirical tests. In this chapter we will refer to it as TPE.

204 | analysis of core theories

Conceptualization of Media

TPE’s conceptualization of media is very broad. It is not limited to any one medium or vehicle. Its scope also seems to extend beyond the media, that is, nonmediated messages could also lead to the dual set of beliefs. In addition, while most of the research focuses on news and information type media messages, TPE also applies to purely persuasive type messages such as advertising and to the persuasive nature of entertainment messages.

Conceptualization of Media Effect

In his introduction of TPE, Davison (1983) clearly positioned it as a media effect where “people will tend to overestimate the influence that mass communications have on the attitudes and behavior of others. More specifically, individuals who are members of an audience that is exposed to a persuasive communication (whether or not this communication is intended to be persuasive) will expect the communi- cation to have a greater effect on others than on themselves” (Davison, 1983, p. 3).

It is unclear where Davison placed the agency of the effect. There is no evi- dence that Davison or subsequent scholars regard TPE as an acquisition type effect where people are exposed to particular beliefs that they acquire as is. Instead, it appears that the media messages act more like a trigger of cognitive processes that result in formulating beliefs about oneself and others. If that is the case, then this would indicate that triggering is the type of media effect. But there is more required in order to have a TPE and that is a difference between first-person beliefs and third-person beliefs. The core essence of TPE is the difference, and it appears that the more active influence for this difference is not from the media but from cognitive processes within the individual.

Conceptualization of Media Influence

Scholars who have tested this effect seem to regard exposure to media messages as an essential antecedent to the effect, although few test for media exposure and use that as a variable to predict the effect. Instead, almost all tests of TPE focus on a variety of cognitive processes within individuals in order to explain the effect.

The conceptualization of media influence is slightly different depending on the methodologies used by researchers. Researchers who use experiments to test TPE typically examine the influence of particular message characteristics by design- ing different treatments that vary those characteristics, so that they can identify

third-person theory | 205

which characteristics in media messages are most related to TPE. Researchers who use surveys necessarily take a more general approach to influence, that is, they do not examine particular message characteristics but instead assume that the media exposures have already triggered the cognitive processes required to establish the different beliefs.

Original Components

Key Concepts

The two original concepts in TPE were third-person beliefs and first-person beliefs. These are sometimes referred to as perceptions, attitudes, and even behav- iors, but they are beliefs because they fit the definition of being cognitions about the probability that an object or event is associated with a given attribute (Fishbein & Ajzen, 1975), that is, faith that something is real or is true. They are not per- ceptions in the sense that they are descriptions of what humans take in through the five senses; they are not evaluations that are required for attitudes; and they are not behaviors because they are typically not performed, although some surveys use self-reports of patterns of past actions or intentions of later actions, which should be considered beliefs rather than actual behaviors.

A third-person belief is a cognition that an individual holds about the per- suasibility of other people, that is, the media have a strong influence on other people. In contrast, a first-person belief is a cognition an individual holds about the persuasibility of one’s self. Thus the two belief concepts of TPE both focused on the idea of persuasibility, that is, the degree to which people could be (and are) influenced by exposure to media messages.

Core Propositions

The central proposition of TPE predicts a contrast between third-person beliefs and first-person beliefs, that is, there must be a difference in the two beliefs with the third-person belief indicating a higher susceptibility to media influence than a first-person belief. When individuals are exposed to a persuasive message (whether or not this communication is intended to be persuasive) in the media, they say they believe that message will have a greater effect on others than on themselves (Davison, 1983, p. 3).

206 | analysis of core theories

Empirical Testing

The explanatory framework for the TPE has expanded considerably since the effect was first described by Davison (1983). This expansion through testing the main idea and looking for cognitive processes to explain the TPE is what has enabled the original simple idea to grow into a system of explanation.

Stimulating Scholarly Attention

TPE has attracted the attention of scholars starting with its introduction and the literature continues to grow. By 1999, TPE had already created a literature of 45 published articles (Perloff, 1999). A year later Paul, Salwen, and Dupagne (2000) conducted a meta-analysis of this literature, which was based on 32 empirical stud- ies of TPE with 121 separate effect sizes reported. Then a few years later Sun et al. (2008) identified 60 papers and publications that tested TPE.

Patterns in the Research

Scope. A good deal of research has been conducted to test the scope of TPE. Table 9.1 shows that the TPE has been tested across a variety of messages, coun- tries, media, and types of people as characterized by demographics. To date, researchers have yet to identify areas where TPE fails to find support.

Outcome variables. The primary outcome variable in TPE has been belief. Although some scholars have referred to the outcome variable as perceptions of others and self, those studies are really testing beliefs (mental constructions of what exists) rather than perceptions (interpretations of stimuli experienced through humans’ five senses).

Over time, researchers have also made claims that they are testing TPE with behaviors. Scholars who have found support for the behavioral component have generally operationalized behavior as a willingness to censor content to stop the content from having the perceived negative persuasive impact on others (Perloff, 1999). Specifically, scholars have demonstrated that TPP predicts willingness to censor pornography (Gunther, 1995), television violence (Hoffner et al., 1999; Rojas, Shah, & Faber, 1996; Salwen & Dupagne, 1999), cigarette, beer, liquor, and gambling advertising (Shah et al., 1999), and antisocial rap music (McLeod et al., 1997). However, researchers have not found that third-person belief pre- dicts willingness to censor news or political media content including censorship of press coverage of the O.J. Simpson trial (Salwen & Driscoll, 1997), support

third-person theory | 207

for an independent commission to regulate political communication or censor- ship (Rucinski & Salmon, 1990), or a Holocaust-denial advertisement (Price et al., 1998).

When reviewing this literature, Perloff (1999) pointed out that these stud- ies do not measure actual behavior; instead they measure behavioral intentions. Because we know there is very little correlation between what people say they intend to do in surveys and their actual behaviors, these outcome variables should be considered as beliefs instead of behaviors.

Table 9.1. Empirical Patterns in Testing of TPE

Testing for Scope

Across types of messages General news coverage (Mutz, 1989; Perloff, 1989; Salwen & Driscoll, 1997)

Of political campaigns (Salwen, 1998) Of environmental problems ( Jensen & Hurley, 2005; Tewksbury et al., 2004) Of public opinion polls (Pan, Abisaid, Paek, Sun, & Houden, 2006; Price & Stroud,

2006) Of health issues (Weinstein, 1980)

Persuasive messages C ommercial advertising (Gunther & Thorson, 1992; Henriksen & Flora, 1999; Huh

et al., 2004; Meirick, 2005; Shah et al., 1999), Political ads (Cohen & Davis, 1991; Meirick, 2004; Rucinski & Salmon, 1990), Public relations messages (Park & Salmon, 2005) Public Service Announcements (PSAs) (Duck & Mullin, 1995)

Entertainment type messages V iolence (Duck & Mullin, 1995; Hoffner et al., 1999; Rojas et al., 1996 ; Salwen &

Dupagne, 1999) P ornography (Gunther, 1995; Lee & Tamborini, 2005; McLeod et al., 1997; Salwen

& Dupagne, 1999; Scharrer, 2002) Rap music (McLeod et al., 1997)

Across countries United States (most of the above-mentioned studies) Germany (Peiser & Peter, 2000) Israel (Tsfati & Cohen, 2005) Australia (Duck & Mullin, 1995; Innes & Zeitz, 1988)

Across media (Antonopoulos, Veglis, Gardikiotis, Kotsakis, & Kalliris, 2015) Across demographic groupings of age, education, and gender (Andsager & White, 2007;

Henriksen & Flora, 1999; Paul et al., 2000; Tiedge et al., 1991)

208 | analysis of core theories

Explanations. A good deal of research was conducted to try to identify factors of influence that could explain why TPE occurs (see Table 9.2). These factors of influence could be arranged into three groups: factors about the media messages, factors about audience members, and processes.

In his most recent review of the TPE literature Perloff (2009) lists nine fac- tors that have been found useful in explaining the TPE. These factors are: self-en- hancement, need for control, projection, attributional biases, focus of attention, media schemas, perceived media exposure, self-categorization, and lack of access to own mental processes. The most important of these appear to be desirability of media messages, need for self-enhancement, and social distance.

Desirability of media message. Perhaps the most important contribution in the TPE empirical literature has been the finding that the desirability of a media message is a crucial factor in explaining the effect. Message desirability refers to the distinction between whether media messages are perceived as antisocial or prosocial.

This effect was first noticed by Innes and Zeitz (1988) who found that their research participants who were exposed to content with a violent message exhib- ited traditional TPEs but that other participants who were exposed to a public service announcement exhibited the reverse. They described this reverse effect, however, only as “something akin to a third person effect” (p. 461). Several years later, Cohen and Davis, who found that people tended to overestimate the effect of attack advertisements for disliked candidates on themselves than on others,

Table 9.2. Factors and Processes Found to Enhance TPE

Factors about People Desire for self enhancement (Hoornes & Ruiter, 1996) S ocial distance between the self and comparison groups (Cohen & Davis, 1991;

Cohen et al., 1988; Davison, 1983; Duck et al., 1995; Eveland et al., 1999; Gunther, 1991; McLeod et al., 1997, 2001; Meirick, 2004, 2005; Scharrer, 2002; White, 1997)

Reference groups (Meirick, 2004) P erceived desirability of messages (Chapin, 2000; Gunther & Mundy, 1993; Gunther

& Thorson, 1992; Hoorens & Ruiter, 1996; Innes & Zeitz, 1988; White, 1997) Factors about Media Messages

Susceptibility and severity (Shah et al., 1999) Emotion eliciting characteristics of ads and PSAs (Gunther & Thorson, 1992)

Processes Social comparison process (Park & Salmon, 2005) Attributional error (Gunther, 1991, 1995; Rucinski & Salmon, 1990)

third-person theory | 209

coined the term “reverse third-person effect” (1991, p. 687). The same year, Tiedge et al. (1991) coined the term “first-person effect” to refer to the perceived effects of media on self as being more than on others.

Most TPE studies, especially early on, dealt with messages that would be seen as “not smart to be influenced by” (Gunther & Mundy, 1993); however, it was found that the effect diminished or reversed into a first-person effect when the media message in question was prosocial or could have desirable consequences. Therefore the type of message (prosocial or antisocial) was found to be an essential explantor of TPE. With prosocial content, we should expect people to say they are strongly influenced by those messages and even more so than other people; this is known as the first-person effect.

Researchers have continually found that what was known as the TPE was strongest on beliefs about media influence of antisocial messages, especially violent and hateful messages (Andsager & White, 2007). However, when people were asked to consider the influence of media’s prosocial messages, the effect was the opposite (Chapin, 2000; Gunther & Mundy, 1993; Gunther & Thorson, 1992; Hoorens & Ruiter, 1996; Innes & Zeitz, 1988; White, 1997). To illustrate, researchers found that when people were asked about media influence on antisocial topics, such as violence, irresponsible sex, and illegal drug use, respondents were likely to say that other people were much more influenced by those media portrayals than they were, which is the pattern of the TPE. However, when respondents were asked about the media influence of prosocial messages that portrayed positive traits such as altruism and generosity, those respondents were likely to say that they themselves were more influenced than other people. This later pattern has been labeled the “first-person effect.”

A meta-analysis of studies of TPP found that message desirability was the most important predictor of TPP (Sun et al., 2008). TPEs are particularly pro- nounced when the message is perceived as undesirable—that is, when people infer that “this message may not be so good for me” or “it’s not cool to admit you’re influenced by this media program.” In line with these predictions, people have been found to perceive content that is typically thought to be antisocial to have a larger impact on others than on themselves (e.g., television violence, pornogra- phy, antisocial rap music) (Perloff, 2009). Indeed, many researchers have found evidence that undesirable messages, such as violent and hateful messages, yield a greater TPE (Andsager & White, 2007; Duck & Mullin, 1995; Gunther & Hwa, 1996; Gunther & Thorson, 1992; Hoornes & Ruiter, 1996). On the other hand, when messages are perceived as desirable, people are not so likely to exhibit a TPE. According to Perloff (2009), the first-person effect, or reversed TPE, is more com- mon for desirable messages.

210 | analysis of core theories

Need for self-enhancement. The patterns described above leave us with a ques- tion about why people would hold these beliefs about media influence on them- selves and on other people. Perloff (1999) in his review of the early TPE literature suggested that people are motivated to preserve self-esteem even if they have to maintain unrealistically positive beliefs about themselves, that is, most people believe it is not socially desirable for them to admit they are strongly influenced by media messages, presumably because that would make them appear ignorant or weak. This explanation was supported by studies that tested the differences between desirable messages and undesirable message and found a greater TPE with undesirable messages.

According to the self-enhancement view, if the TPE is driven by a desire to preserve self-esteem, people should be willing to acknowledge effects for commu- nications that are regarded as socially desirable, healthy, or otherwise good for the self (Hoornes & Ruiter, 1996; Tal-Or et al., 2009).

Social distance. In his introduction of TPE, Davison (1983) suggested that social distance was likely an explanation for the effect. Thus we should expect the pattern of TPE to be stronger when respondents perceive a greater distance between themselves and the others. That is, the more respondents think that a par- ticular group of people are different from themselves, the stronger will be the TPE. “The greater the perceived social distance tween self and others, the easier it is to assume the third persons will fall prey to the effects that ‘I’ see through” (p. 364). The disparity of self and other is increased as perceived distance between self and comparison others is increased (Meirick, 2004, 2005).

Perloff (1999) said that “the nature of the social distance comparison between self and other depends in important ways on the identity of the hypothetical oth- ers” (p. 363). According to this notion, the magnitude of the TPE increases as the social distance between self and comparison others increases, or the hypothetical others are defined in larger, broader terms.

The social distance explanation has been tested often and typically found to hold (Cohen & Davis, 1991; Cohen, Mutz, Price, & Gunther, 1988; Duck, Hogg, & Terry, 1995; Gunther, 1991). Although social distance is not a necessary con- dition for the TPE to occur, increasing the social distance makes the TPE larger. In their meta-analysis, Andsager and White (2007) concluded that “Research consistently finds that others who are anchored to self as a point of reference are perceived to be less influenced by persuasive messages than are others who are not defined and, therefore, not anchored to any point of reference at all” (p. 92).

Target groups. The social distance explanation was challenged by findings that perceived exposure of a group to a message was a better predictor of perceived effects than perceived similarity to the group (Eveland, Nathanson, Detenber,

third-person theory | 211

& McLeod, 1999; McLeod, Detenber, & Eveland, 2001; McLeod et al., 1997; Meirick, 2004, 2005). Thus it was reasoned that TPE was influenced to the extent that if a group is perceived as the target of a type of media content, perceived effects on them will be greater. Others have discussed this target explanation but mainly in terms of its tenability as an alternate explanation for their findings (Meirick, 2004; Scharrer, 2002). Despite the strong results of the target explanation cited above, the effect has been tested only for antisocial music lyrics. It has not been tested for messages where there would likely be less perceived variance in exposure, nor has it been tested for desirable messages. There is some evidence to suggest that the target explanation might not hold for positive messages (Meirick, 2004; Scharrer, 2002) perhaps because of incompatibility between beliefs about a group’s exposure and (heretofore underexamined) beliefs about a group’s predispositions toward the message’s advocated behavior.

Other explanations. Researchers have also tested other explanations for TPE. Perloff (1999) notes that the majority of TPE studies attribute the psychologi- cal underpinnings of the effect to either attribution theory or biased optimism. Attribution theory predicts that actors tend to attribute their actions to situational factors, while observers tend to attribute the same actions to dispositional fac- tors. For example, attribution theory predicts that a student who turns in a late assignment may explain to the professor that the tardiness is uncharacteristic and due to a situational factor like an unusual computer problem, while the professor might believe the tardiness was due instead to a dispositional factor like the stu- dent’s laziness. In the context of the TPE, then, attribution theory explains why a person may think that he or she understands the underlying persuasive aspects of the message, while others’ dispositional flaws prevent them from perceiving those same aspects (Perloff, 1999).

Biased optimism predicts that people tend to judge themselves as less likely than others to experience negative consequences and, conversely, that people tend to judge themselves as more likely than others to experience positive events. In the context of the TPE hypothesis, biased optimism explains why people judge them- selves as being less likely than others to be affected by persuasion (Perloff, 1999).

Empirical Validity

Reviews of this growing literature typically conclude that there is strong and con- tinuing support for TPE in general but that there may be particular areas where the theory is weak or lacking support. In his review of TPE, Lasorsa (1992) said that about half of the members of any sample show a TPE while the others do not. This is typically enough to show overall support for the theory. However,

212 | analysis of core theories

this finding has led researchers to look for contingent factors to try explaining why some people are susceptible while others are not. Also, researchers who have designed experiments can look for contingencies on the way the topic is presented in stimulus materials.

There have been two meta-analyses of the TPE literature and these both conclude the literature shows strong support for TPE (relative to the support of other theories and other media effects) but that there are some important contin- gencies. Paul et al. (2000) conducted a meta-analysis of 121 separate effect sizes reported and converted these varied statistical indicators into a common metric of an r where positive r indicated a TPE. The findings from their meta-analysis indicate the perceptual component of the TPE hypothesis received robust support (r = .50), especially compared to meta-analyses of other media effects theories. They also tested eight factors to see how stable the effect was: source, method, sampling, respondent, country, desirability, medium, and message. Source referred to published and unpublished studies. Method referred to surveys or experiments. Sampling referred to random and nonrandom. Respondent was either college stu- dents or noncollege people. Country was U.S. and other countries. Desirability had three values—desirable to believe, undesirable, and neither. Medium had four values (media in general, television radio, newspapers, and other). Message variable had seven values (message in general, pornography, television violence, commercial ads, politics, nonpolitical news, and other). They report finding three of these eight to be most significant: sampling, respondent, and message. Samples that were nonrandom yielded greater TPE differences than samples obtained from random samples. As for type of respondent, they found that samples composed of stu- dents yielded greater TPE differences than nonstudent samples. As for message, different types of content (e.g., general media messages, pornography, television violence, commercial advertisements, political content, nonpolitical news) have differing effects on the size of the obtained TPPs.

A few years later, Sun et al. (2008) conducted a meta-analysis of 106 studies that reported 372 effect sizes. They tested the TPE for stability using 16 test fac- tors. Of these, seven were research characteristics (study setting, data collection method, design, study population, sampling method, measure for perceived effects, and domain of perceived effects). Four were message characteristics (desirability of message, message topic domain, persuasive intent, and functional focus). The rest were referent characteristics, such as geographical distance and sociodemograph- ics. After finding an average effect size estimate of d = 0.646 (r = .307), the authors said, “The effect size, however, is significantly smaller than that reported in Paul et al.’s (2000) study, which, as we demonstrated, contained serious overestimation because they did not (a) apply the right effect size formula for within-subjects

third-person theory | 213

designs nor (b) address the statistical dependency among multiple effect sizes extracted from the same study” (p. 294).

Sun et al. (2008) also found the influence was not sensitive to five tested factors (research setting, population, between- or within-subjects design, model of data collection, or whether single- or multiple-item measures were used) but that the other 12 factors they tested were all found to be significant with message desirability being the most important. In summary, they reported “Results from a series of multilevel models show that the third-person perception is robust and not influenced by variations in research procedures. Desirability of presumed message influence, vulnerability of referent others, referent others depicted as being similar to self, and others being likely audience of the media content in question are sig- nificant moderators” (p. 280).

Theory Development

Conceptual Development

Conceptual criticism. Some reviewers of this research make a distinction between perceptions and behaviors. For example, Tal-Or et al. (2009) argue that “most research on the third-person effect to date has examined this component (perceptual) . . . Yet the reason for Davison’s enthusiasm about TPPs was that he recognized the enormous ways in which these perceptions might impact the real world. However, only in recent years have scholars turned their attention to the behavioral consequences of people’s biased perceptions of media coverage” (p. 104). While this at first sounds like a useful distinction—perceptions vs. behaviors—it is faulty, because the perceptions are really beliefs and when we look at what is regarded at behavioral measures in the published literature, we see that these mea- sures are self-reports of intended behaviors or beliefs about behaviors. For exam- ple, Tewksbury, Moy, and Weis (2004) found that after the media reported on the so-called millennium buy (Y2K) that could cause havoc on computers, people said they were less likely to stockpile supplies (food, water, gasoline, cash, etc.) than other people. Research has also found that after the media make predictions of likely earthquakes, people report they are less likely to prepare their homes for such a disaster than other people. These results are highly suspect as reflecting behaviors; we have known for a long time that what people report as their behaviors and what their actual behaviors are has a weak association at best. It is more likely that these results reflect the desire of people to be consistent in their beliefs, that is, if they say they are less affected by the media then when asked about their behaviors they will say their behaviors exhibit less effect than the presumed behaviors of other people.

214 | analysis of core theories

Conceptual alterations. Over the years, scholars have argued for other proposi- tions to explain TPE by expanding on the original ideas from Davison. These are typically referred to as “corollaries.” There are three such corollaries referred to in the literature as explanations for why the TPE occurs. These are negative influence corollary, social distance corollary, and the target corollary.

Negative influence corollary. This corollary recognizes that the TPE is sensitive to the type of content. If the type of content is antisocial (beliefs about the effects of violence, pornography, drugs, etc.), then the TPE predictions hold, that is, peo- ple believe that others are more influenced by media coverage than they are per- sonally. However, when the content is prosocial, then the opposite is the case, that is, people generally believe that they are more likely to be influenced in a positive way compared to other people. This has sometimes been called the first-person effect, which says that sometimes people will have beliefs that media influence is stronger on them than on other people.

Gunther and Storey (2003) write about “a negative influence corollary” as a contingent factor that limits the general nature of TPE. They argued that “Although the evidence for FPP is not robust in part due to a much smaller num- ber of effect sizes in the extant literature, it is clear that ‘the negative influence cor- ollary’ (Gunther & Storey, 2003) specifies theoretical boundaries of TPP” (p. 294). However, other scholars recognize that this corollary can be incorporated into the theory if it is expanded to encompass all of the presumed media influence (Tal-Or et al., 2009).

Social distance corollary. Another cognitive process that has been associated with TPE has been called the social distance corollary where people choose to dissociate themselves from the others who may be influenced. Perloff (1999) said that “the nature of the social distance comparison between self and other depends in important ways on the identity of the hypothetical others” (p. 363) and this has “come to be called the social-distance corollary. According to this notion, the magnitude of the TPE increases as the social distance between self and compari- son others increases, or the hypothetical others are defined in larger, broader terms. The greater the perceived social distance between self and others, the easier it is to assume the third persons will fall prey to the effects that ‘I’ see through” (p. 364). This effect has been widely cited in the research literature (Tal-Or et al., 2009) and it has been found to have strong empirical support (Paul et al., 2000).

Target corollary. The target corollary predicts that TPE will be stronger on targets who are perceived as being more involved in the issues that are being exam- ined. There is some evidence to suggest that the target corollary might not hold for positive messages (Meirick, 2004), perhaps because of incompatibility between beliefs about a group’s exposure and (heretofore underexamined) beliefs about a

third-person theory | 215

group’s predispositions toward the message’s advocated behavior. Thus was born the “target corollary”—to the extent that a group is perceived as the target of a type of media content, perceived effects on them will be greater.

Methodological Development

Most of the criticism about TPE has focused on methodological issues, especially concerning problems of validity of findings due to the way questions are asked.

Methods used. As for methods patterns, most of these studies are cross-sec- tional surveys although several measured at least two points in time (Price et al., 1998). Also most have used convenience samples (e.g., McLeod et al., 1997; Meirick, 2004, 2005), but a few have used national probability samples (Gunther, 1995; Huh et al., 2004; Salwen & Driscoll, 1997) or more local probability samples (Tiedge et al., 1991). Some have used personal interviews (Duck & Mullin, 1995; Innes & Zeitz, 1988; Shah et al., 1999). And a few have used experiments (Cohen & Davis, 1991; Duck & Mullin, 1995; Gunther & Thorson, 1992; Hoornes & Ruiter, 1996; Park & Salmon, 2005). Many have used undergraduate college stu- dents (Cohen & Davis, 1991; Duck & Mullin, 1995; Gunther & Thorson, 1992; Hoornes & Ruiter, 1996; McLeod et al., 1997; Meirick, 2004, 2005; Park & Salmon, 2005; Price et al., 1998).

Criticism. There has been a continuing disconnect between what Davison originally conceptualized the effect to be and how it has been tested—including tests by Davison himself. In his introduction of TPE, Davison (1983) defined the effect as “people will tend to overestimate the influence that mass communications have on the attitudes and behavior of others” (p. 3). In order to use this claim as it is worded requires the identification of a true level of influence to use as a standard upon which to determine if people are underestimating or overestimating media influence. Nowhere in the literature does a scholar tackle this task, perhaps because it is a near impossible thing to do. However, we could make a minor alteration in operationalizing tests and use as a standard the actual risk levels for various things, such as chances of being in an automobile accident due to drinking, drug use, or mobile device use. Then ask people to estimate the levels of such risks on other people. So if they estimate a risk of something covered in a media message to be 10 % and the actual level of risk to be 2 %, then this would be evidence of an overestimation. However, this kind of test exists nowhere in the TPP literature. Instead, the typical test is to ask people to estimate their own risk as well as the risk on others; if the estimate of the third person is higher than the estimate on self then this is taken as adequate evidence of a TPE. And if the estimate of the third person is lower than the estimate on self then this is taken as adequate evidence of

216 | analysis of core theories

a first-person effect. Therefore, in the empirical literature, the standards of com- parison are always beliefs about media influence or risk—never actual levels of influence or risk.

A second criticism is that evidence for TPE might be a measurement arti- fact if it is found that the first question asked in a study serves as an anchor for respondents in answering the second question (Perloff, 1999). Many tests have been run that have altered question orders and formats; these tests have all found support for TPE (Brosius & Engel, 1996; David, Lui, & Myser, 2004; Gunther, 1995; Gunther & Hwa, 1996; Price & Tewksbury, 1996; Salwen & Driscoll, 1997; Tiedge et al., 1991). For example, Price and Tewksbury tested whether the TPE was a methodological artifact as a result of asking participants self-other ques- tions in close proximity. Using a three-condition experiment in which they asked participants in the first condition self-only questions, participants in the second condition other-only questions, and participants in the third condition self and other questions, Price and Tewksbury’s (1996) results indicate consistent estimates of self and other estimates across conditions. These results, then, indicate the effect is not the result of a methodological artifact.

Current Challenges

There seem to be two current challenges facing the theory. First, there are lingering questions about whether it qualifies as a theory. Second, it needs to increase its level of precision.

Qualify as Theory

Scholars who write about TPE sometimes refer to it only as an effect, sometimes as an hypothesis, and sometimes as a theory. When Davison (1983) first intro- duced the idea he called it an effect then ran several tests calling it the “third-per- son effect hypothesis.” In his major review of TPE, Perloff (2009) refers to it as an effect, a hypothesis, and a theory by using all three terms. He calls it an effect in many places, most prominently in the title “Mass media, social perception, and the third-person effect.” He also says “the third-person effect is more hypothesis than full-blown theory” (p. 254) but does not explain what is missing in order for him to consider it a theory. But then he also said that TPE is “ranked as the fifth most popular theory in 21st century mass communication research” (pp. 252–253).

It is understandable why this ambivalence existed when Davison first intro- duced the TPE, but over time the contributions that scholars have made to this

third-person theory | 217

literature have added components such that it now meets the definition of theory as derived in Chapter 2. TPE has evolved into a system of explanation that features key concepts arranged in propositions that express the relationships between two or more of those concepts. Each of those propositions is testable and they have been assembled into a set that has internal coherence as a system of explaining how the media along with other factors work together to influence the occurrence of a media effect. It also has stimulated numerous empirical tests of those propositions. For these reasons, it can—and should—be regarded as a theory.

TPE, however, is different from most of the other theories in one import- ant respect, and this will present a considerable challenge going forward. Most of the theories analyzed in this book were introduced by a media effects scholar who continued working on it by testing it and publishing reviews of the grow- ing empirical literature over several decades. In contrast, TPE was introduced by a sociologist who did not continue publishing tests of this theory and who did not publish reviews of the empirical literature testing his theory. However, other scholars—most notably David Perloff, Al Gunther, Michael Salwen, and Vincent Price—have each published several tests, provided probing analyses of the theory, and worked to add to its scope and explanatory power. Thus for the purposes of the analysis in this book, it has been less useful to talk about the original concep- tualization of Davison, which was highly speculative and more useful to regard the theory as the set of all work that has been done over the years. But moving forward this presents a challenge about who will assume the role of guiding future development and arbitrating disagreements that might arise over how the theory should be altered and elaborated.

Whoever steps up to guide the continued development of TPE will be able to continue to mine ideas from the larger media effects literature. Over time, researchers have drawn on a variety of psychological theories to justify the TPE. Few, however, explicitly linked these theories to the TPE. Some researchers have used ego involve- ment (e.g., Perloff, 1989; Vallone, Ross, & Lepper, 1985), the elaboration likelihood model (e.g., White, 1997), and social categorization theory, but most have relied on attribution theory (e.g., Gunther, 1991; Hoffner et al., 1999; Rucinski & Salmon, 1990) and biased optimism (e.g., Brosius & Engel, 1996; Gunther & Mundy, 1993; Rucinski & Salmon, 1990) to explain the theoretical underpinnings of the TPE.

Precision

This theory struggled from the beginning and still struggles with precision in the defining and labeling of concepts. This can be seen in the selection of a name for the theory. When Davison (1983) introduced this theory into the scholarly

218 | analysis of core theories

literature, he used the title “The third-person effect in communication” and contin- ually referred to it as the TPE. But he could have just as well called it the first-per- son effect. Better yet it should be regarded as both, but that would be too awkward.

A second area where precision is a major problem is with the conception of behavior. Reviewers of the literature make claims about the theory having two stages: perception and behavior. But the measures of behavior have tapped into beliefs (self-reports and intentions) rather than actual actions. If the theory is interested in explaining behavior, then researchers will need to design experiments that move beyond a reliance on questionable self-reports and test for actual behav- iors (Tal-Or et al., 2009).

Conclusions

What started out as a report of an intriguing effect in 1983 has since grown broader in scope and as a system of explanation until it is now regarded as Presumed Media Influence (Tal-Or et al., 2009). The core idea has received considerable support over the years, and researchers have sought to think up then test explanations for why TPE occurs so consistently. We now have a growing system of explanation that has a solid core of explanation about TPE and first-person effects that has been well supported by empirical tests. In addition, the theory has been elaborated with three corollaries beyond the core and is likely to continue growing its explan- atory power in the future.

Key Sources

Origin Davison, W. P. (1983). The third-person effect in communication. Public Opinion Quarterly,

47(1), 1–15. doi:10.1086/268763.

Reviews Lasorsa, D. L. (1992). How media affect policymakers: The third-person effect. In J. D.

Kennamer (Ed.), Public opinion, the press and public policy (pp. 163–175). New York, NY: Praeger.

Paul, B., Salwen, M. B., & Dupagne, M. (2000). The third-person effect: A meta-analysis of the perceptual hypothesis. Mass Communication & Society, 3(1), 57–85. doi:10.1207/ s15327825mcs0301_04

Perloff, R. M. (1999). The third-person effect: A critical review and synthesis. Media Psychology, 1(4), 353–378. doi:10.1207/s1532785xmep0104_4.

third-person theory | 219

References

Andsager, J. L., & White, H. A. (2007). Self versus others: Media, messages, and the third-person effect. Mahwah, NJ: Erlbaum. ISBN 9780805857160.

Antonopoulos, N., Veglis, A., Gardikiotis A., Kotsakis, R., & Kalliris, G. (2015). Web Third- person effect in structural aspects of the information on media websites. Computers in Human Behavior, 44(3), 48–58. doi:10.1016/j.chb.2014.11.022

Brosius, H-B., & Engel, D. (1996). The causes of third-person effects: Unrealistic optimism, impersonal impact, or generalized negative attitudes towards media influence? International Journal of Public Opinion Research 7, 142–162.

Chapin, J. R. (2000). Third-person perception and optimistic bias among urban minority at-risk youth. Communication Research, 27, 51–81.

Cohen, J., & Davis, R. G. (1991). Third-person effects and the differential impact in negative polit- ical advertising. Journalism Quarterly, 68(4), 680–688. doi:10.1177/107769909106800409

Cohen, J., Mutz, D., Price, V., & Gunther, A. C. (1988). Perceived impact of defamation. Public Opinion Quarterly, 52, 161–173.

David, P., Liu, K., & Myser, M. (2004). Methodological artifact or persistent bias? Testing the robustness of the thirdperson and reverse thirdperson effects for alcohol messages. Communication Research, 31, 206–233.

Davison, W. P. (1983). The third-person effect in communication. Public Opinion Quarterly, 47(1), 1–15. doi:10.1086/268763

Perloff, R. M. (2009). Mass media, social perception, and the third-person effect. In J. Bryant & M. B. Oliver (Eds.), Media effects: Advances in theory and research (3rd ed., pp. 252–268). London: Routledge. ISBN 9781135591106.

Sun, Y., Pan, Z., & Shen, L. (2008). Understanding the third-person percep- tion: Evidence from a meta-analysis. Journal of Communication, 58(2), 280–300. doi:10.1111/j.1460-2466.2008.00385.x.

Tal-Or, N., Tsafati, Y., & Gunther, A. C. (2009). The influence of presumed media influ- ence: Origins and implications of the third-person perception. In R. L. Nabi & M. B. Oliver (Eds.), Media processes and effects (pp. 99–112). Los Angeles, CA: Sage.

Criticisms Perloff, R. M. (1999). The third-person effect: A critical review and synthesis. Media

Psychology, 1(4), 353–378. doi:10.1207/s1532785xmep0104_4. Tal-Or, N., Tsafati, Y., & Gunther, A. C. (2009). The influence of presumed media influ-

ence: Origins and implications of the third-person perception. In R. L. Nabi & M. B. Oliver (Eds.), Media processes and effects (pp. 99–112). Los Angeles, CA: Sage.

220 | analysis of core theories

Duck, J. M., Hogg, M. A. & Terry, D. J. (1995). Me, us and them: Political communication and the third person effect in the Australian Federation Election. European Journal of Social Psychology, 25, 195–215.

Duck, J. M., & Mullin, B.-A. (1995). The perceived impact of the mass media: Reconsidering the third person effect. European Journal of Social Psychology, 25(1), 77–93. doi:10.1002/ ejsp.2420250107

Eveland, W. P., Jr., Nathanson, A. I., Detenber, B. H., & McLeod, D. M. (1999). Rethinking the social distance corollary. Perceived likelihood of exposure and the third-person perception. Communication Research, 26, 275–302.

Fishbein, M. & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.

Gunther, A. C. (1991). What we think others think: Cause and consequence in the third-person effect. Communication Research, 18, 355–372.

Gunther, A. C. (1995). Overrating the X-rating: The third-person perception and sup- port for censorship of pornography. Journal of Communication 45(1), 27–38. doi:10.1111/j.1460-2466.1995.tb00712.x.

Gunther, A. C. & Hwa, A. P. (1996). Public perceptions of television influence and opinions about censorship in Singapore. International Journal of Public Opinion Research, 8(3), 248– 265. doi:10.1093/ijpor/8.3.248

Gunther, A. C., & Mundy, P. (1993). Biased optimism and the third-person effect. Journalism Quarterly, 70, 58–67.

Gunther, A. C. & Storey, J. D. (2003). The influence of presumed influence. Journal of Communication, 53, 199–215.

Gunther, A. C., & Thorson, E. (1992). Perceived persuasive effects of commercials and public service announcements: The third-person effect in new domains. Communication Research, 19(5), 574–596. doi:10.1177/009365092019005002

Henriksen, L., & Flora, J.A. (1999). Third-person perception and children: Perceived impact of pro- and anti-smoking ads. Communication Research, 26, 643–665.

Hoffner, C., Buchanan, M., Anderson, J. D., Hubbs, L. A., Kamigaki, S. K., Kowalczyk, L., … Silberg, K. J. (1999). Support for censorship of television violence: The role of the third-per- son effect and news exposure. Communication Research, 26, 726–742.

Hoornes, V., & Ruiter, S. (1996). The optimal impact phenomenon: Beyond the third person effect. European Journal of Social Psychology 26(4), 599–610. doi:10.1002/ (sici)1099-0992(199607)26:4<599::aid-ejsp773>3.0.co;2-7

Huh, J., Delorme, D., & Reid, L.N. (2004). The third-person effect and its influence on behavioral outcomes in a product advertising context: The case of direct-to-consumer prescription drug advertising. Communication Research 31(5): 568–599. doi:10.1177/0093650204267934

Innes, J. M., & Zeitz, H. (1988). The public’s view of the impact of the mass media: A test of the “third-person” effect. European Journal of Social Psychology 18(5), 457–463. doi:10.1002/ ejsp.2420180507

third-person theory | 221

Jensen, J. D., & Hurley, R. J. (2005). Third-person effects and the environment: Social distance, social desirability, and presumed behavior. Journal of Communication, 55, 242–256.

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

McLeod, D. M., Detenber, B. H., & Eveland, W. P., Jr. (2001). Behind the third-person effect: Differentiating perceptual processes for self and others. Journal of Communication, 51, 678–695.

McLeod, D. M., Eveland, W. P., & Nathanson, A. I. (1997). Support for censorship of violent and misogynic rap lyrics: An analysis of the third-person effect. Communication Research 24(2), 153–174. doi:10.1177/009365097024002003

Meirick, P. C. (2004). Topicrelevant reference groups and dimensions of distance: political advertising and first and thirdperson effects. Communication Research, 31, 234–255.

Meirick, P. C. (2005). Rethinking the target corollary: The effects of social distance, perceived exposure, and perceived predispositions on first-person and third-person perceptions. Communication Research 32(6): 822–843. doi:10.1177/0093650205281059

Mutz, D. (1989). The influence of perception of media influence. International Journal of Public Opinion Research, 1, 3–24.

Pan, Z., Abisaid, J. L., Paek, H-J., Sun, Y., & Houden, D. (2006). Exploring the perceptual gap in perceived effects of media repots of opinion polls. International Journal of Public Opinion Research, 18, 340–350.

Park, H. S., & Salmon, C. T. (2005). A test of the third-person effect in public rela- tions: Application of social comparison theory. Journalism Quarterly 82(1), 25–43. doi:10.1177/107769900508200103

Paul, B., Salwen, M. B., & Dupagne, M. (2000). The third-person effect: A meta-analysis of the perceptual hypothesis. Mass Communication & Society, 3(1), 57–85. doi:10.1207/ s15327825mcs0301_04

Peiser, W., & Peter, J. (2001). Explaining individual differences in third-person perception: A limits/possibilities perspective. Communication Research, 28, 156–180.

Perloff, R. M. (1989). Ego-involvement and the third-person effect of televised news coverage. Communication Research, 16, 236–262.

Price, V., & Stroud, N. J. (2006). Public attitudes toward polls: Evidence from the 2000 U.S. pres- idential election. International Journal of Public Opinion Research, 18, 393–421.

Price, V., & Tewksbury, D. (1996). Measuring the third-person effect of news: The impact of question order, contrast and knowledge. International Journal of Public Opinion Research 8(2), 120–141. doi:10.1093/ijpor/8.2.120

Price, V., Tewksbury, D., & Huang, L-N. (1998). Third-person effects on publica- tion of a holocaust-denial advertisement. Journal of Communication 48(2), 3–26. doi:10.1111/j.1460-2466.1998.tb02745.x

222 | analysis of core theories

Rojas, H., Shah, D. V., & Faber, R. J. (1996). For the good of others: Censorship and the third-person effect. International Journal of Public Opinion Research 8(2): 163–186. doi:10.1093/ijpor/8.2.163

Rucinski, D., & Salmon, C. T. (1990). The “other” as the vulnerable voter: A study of the third-person effect in the 1988 U.S. presidential campaign. International Journal of Public Opinion Research 2(4), 345–368. doi:10.1093/ijpor/2.4.345

Salwen, M. B. (1998). Perception of media influence and support for censorship: The third-per- son effect in the 1996 presidential election. Communication Research, 25, 259–285.

Salwen, M. B., & Dupagne, M. (1999). The third-person effect: Perceptions of the media’s influ- ence and immoral consequences. Communication Research, 26, 523–549.

Salwen, M. B., & Dupagne, M. (2001). Third-person perception of television violence: The role of self-perceived knowledge. Media Psychology, 3, 211–230.

Salwen, M. B., & Driscoll, P. D. (1997). Consequences of third-person perception in support of press restrictions in the O.J. Simpson trial. Journal of Communication 47(2), 60–75. doi:10.1111/j.1460-2466.1997.tb02706.x

Scharrer, E. (2002). Third-person perception and television violence: The role of out-group ste- reotyping in perception of susceptibility to effects. Communication Research, 29, 681–704.

Shah, D. V., Faber, R. J., & Youn, S. (1999). Susceptibility and severity: Perceptual dimen- sions underlying the third-person effect. Communication Research 26(2): 240–267. doi:10.1177/009365099026002006

Sun, Y., Pan, Z., & Shen, L. (2008). Understanding the third-person percep- tion: Evidence from a meta-analysis. Journal of Communication, 58(2), 280–300. doi:10.1111/j.1460-2466.2008.00385.x

Tewksbury, D., Moy, P., & Weis, D. S. (2004) Preparations for Y2K: Revisiting the behavioral components of the third-person effect. Journal of Communication, 54, 138–155.

Tiedge, J. T., Silverblatt, A., Havice, M. J., & Rosenfeld, R. (1991). Discrepancy between per- ceived first person and perceived third-person mass media effects. Journalism Quarterly, 68(1/2), 141–154. doi:10.1177/107769909106800115

Tsfati, Y., & Cohen, J. (2004). Object-subject distance and the third person perception. Media Psychology, 6, 335–361.

Tsfati, Y., & Cohen, J. (2005). The influence of presumed media influence on democratic legiti- macy: The case of Gaza settlers. Communication Research, 32, 794–821.

Vallone, R., Ross, L., & Lepper, M. (1985). The hostile media phenomenon: Biased perception and perceptions of media bias in coverage of the Beirut massacre. Journal of Personality and Social Psychology, 49, 577–585.

Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of Personality and Social Psychology, 39, 806–820.

White, H. A. (1997). Considering interacting factors in the third-person effect: Argument strength and social distance. Journalism & Mass Communication Quarterly, 74, 557–564.