Essay

Journalism & Mass Communication Quarterly 2016, Vol. 93(4) 906 –922

© 2016 AEJMC Reprints and permissions:

sagepub.com/journalsPermissions.nav DOI: 10.1177/1077699016637108

jmcq.sagepub.com

Media Reputation Measurement and Evaluation of Media Companies

The Effects of Media Effects: Third-Person Effects, the Influence of Presumed Media Influence, and Evaluations of Media Companies

Brett Sherrick1

Abstract Prior research in the third-person effects domain has shown that people who believe in harmful media effects are more willing to engage in preventive or accommodative strategies, such as censorship. This research extends that supposition by testing a thus-far unstudied strategy: negative evaluations of media companies. Results show that an overall belief in harmful media effects is connected to negative evaluations of the media companies potentially responsible for those effects. The third-person perceptual gap is not related to these negative evaluations of media companies, suggesting important differences between third-person effects research and influence of presumed media influence research.

Keywords third-person effects, influence of presumed media influence, evaluations of media companies

In the aftermath of the Sandy Hook Elementary School shooting, in which a lone gun- man opened fire at an elementary school in Connecticut, a number of groups and individuals were eager to blame violent media—particularly violent video games—for the shooter’s mentality, despite limited evidence that the shooter consumed violent

1University of Alabama, Tuscaloosa, USA

Corresponding Author: Brett Sherrick, Department of Journalism, University of Alabama, Box 870172, Tuscaloosa, AL 35487, USA. Email: [email protected]

637108 JMQXXX10.1177/1077699016637108Journalism & Mass Communication Quarterly XX(X)Sherrick research-article2016

Sherrick 907

media. Many similar acts of unnecessary violence have been blamed on the harmful effects of violent media. In part, this phenomenon can be explained by third-person effects and related research, which has shown that the general public does believe in media effects—particularly when it comes to effects on others—and will change their attitudes and behaviors accordingly. In fact, the “influence of presumed media influ- ence” has been verified in a number of recent studies (e.g., Gunther & Storey, 2003; Tal-Or, Cohen, Tsfati, & Gunther, 2010; Tsfati & Cohen, 2005). Moreover, third-per- son effects literature shows that a belief in media effects has potential behavioral out- comes, such as attempts to counteract or censor the offending media (Perloff, 2009).

Presumably, individuals who believe in the power of harmful media effects enough to consequently engage in censorship are likely to have negative perceptions of media companies, but this assumption has received little academic attention. In other words, it is not clear to what degree media companies’ reputations suffer from their stakehold- ers’ perceptions about the harmful effects of media. The current research examines that connection through a survey design, examining whether people who believe in the harmful effects of media are likely to evaluate media companies negatively in terms of corporate social responsibility (CSR), willingness to recommend that company (word of mouth), and general satisfaction with that company.

Effects of a Belief in Media Effects

The third-person perception is the belief that others are more likely to be affected by mediated messages than the self. Third-person perceptions have been consistently found for diverse populations and across various forms of media, including news media, advertising, and entertainment media (Perloff, 2009). For example, Salwen (1998), in a nationwide U.S. survey, found that people believed that others’ opinions of political candidates were more influenced by newspapers, radio, and TV than were their own views; Rojas, Shah, and Faber (1996) demonstrated that college students believed that other people were more affected by violent TV and pornography; and Scharrer and Leone (2006, 2008) found that children believe that other children are more influenced by video games. Moreover, meta-analyses of the third-person percep- tion have confirmed the general perceptual hypothesis, with results showing moderate to large overall effects (Paul, Salwen, & Dupagne, 2000; Sun, Pan, & Shen, 2008). Scholars have hypothesized and tested a number of potential causal factors for this perceptual gap; one of the most common explanations posits that individuals believe themselves to be more resistant to media effects than others because they want to see themselves in a positive light when compared with others (Perloff, 2009). The current research first attempts to confirm the third-person perception hypothesis:

H1: Belief in harmful effects from a media company will be greater when consider- ing others than when considering the self.

A related theory, the “influence of presumed media influence,” grew out of third-per- son effects research but describes the effects of a more general belief in media influence.

908 Journalism & Mass Communication Quarterly 93(4)

Rather than focusing on the gap in perceived media effects between self and others, this research examines an overall belief in media effects (Gunther & Storey, 2003), but many of the expectations and findings are similar to those of third-person effects research. Thus far, research on the influence of presumed media influence has primarily focused on news media, examining political attitudes and even voting behaviors as outcomes (e.g., Cohen & Tsfati, 2009; Cohen, Tsfati, & Sheafer, 2008; Tsfati & Cohen, 2005). However, other researchers have shown effects of the presumed influence of reality TV (Cohen & Weimann, 2008), of beauty and fashion magazines (Park, 2005), and of pro- and anti- smoking messages (Gunther, Bolt, Borzekowski, Liebhart, & Dillard, 2006).

Taken together, research on the influence of presumed media influence and the third-person perception suggests that people do generally believe in media effects and that this belief is likely to be stronger in regard to effects on others than in regard to effects on the self. Moreover, both domains of research have shown that belief in media effects has measurable effects on other attitudes and even behaviors. Most com- monly, research has shown that belief in (negative) media effects may lead people to attempt to prevent those effects, typically through supporting or enacting media cen- sorship. More recently, researchers have focused on strategies of accommodation (Gunther et al., 2006; Gunther, Perloff, & Tsfati, 2008), whereby people modify their existing attitudes or behaviors based on their belief in the effects of media. For exam- ple, people may “comply” with ideas from mass media because they believe that oth- ers will be influenced by the media, so they follow the media as well to fit in with social norms and expectations (Perloff, 2009).

Numerous studies have supported the claim that a belief in harmful media effects is positively related to preventive measures such as a willingness to censor media (e.g., McLeod, Eveland, & Nathanson, 1997; Salwen, 1998; Schmierbach, Boyle, Xu, & McLeod, 2011). For example, Rojas et al. (1996) found—in a survey of select college students—that the third-person perception was related to a willingness to censor gen- eral media, violent TV, and pornography. The third-person perception regarding por- nography was also statistically connected to behavioral intentions to censor pornography. Tal-Or et al. (2010, Study 1) found a similar willingness to censor por- nography when considering the influence of presumed media influence, rather than the third-person effect. Unlike many studies before it, the Tal-Or et al. (2010) study used an experimental design in which participants were either informed of positive effects of pornography or negative or neutral effects of pornography. Results showed that those who were informed of the positive effects of pornography were subsequently less likely to support censorship of pornography than those informed of the negative or neutral effects; as expected, this relationship was mediated by a belief in effects of pornography, suggesting that the experimental manipulation was effective and that belief in presumed influence is causally related to willingness to censor. Finally, although pornography is a common topic for this area of research, other studies have shown that belief in harmful media effects is related to a willingness to censor media such as political news and advertisements (Salwen, 1998), rap music (McLeod et al., 1997), video games (Schmierbach et al., 2011; Wolf, 2010), and reality TV (Cohen & Weimann, 2008; Sun, Shen, & Pan, 2008).

Sherrick 909

Although preventive strategies are often equated with a general willingness to cen- sor in third-person effects research, there are potentially a number of different out- comes of belief in harmful media effects (Gunther et al., 2008). Wolf (2010), for example, found that those who believe that violent video games lead to real-world violence were more likely to both support government monitoring of game content and believe that victims of “copycat” violence should be able to sue game manufactur- ers—two decidedly different outcomes. Importantly, these outcomes also have differ- ent targets (the government and individual game manufacturers, respectively). Conceivably, attitudes or behaviors inspired by a belief in media effects may be tar- geted at the government, which should impose legal censorship; media industries writ large, which should police their member companies; media distributors, which should ensure offending media are unavailable to vulnerable populations; parents, who should better monitor their children’s media use; or individual media companies, which should create less objectionable materials. Previous research on the effects of belief in media influence has not been particularly sensitive to these potentially distinct targets.

To begin to investigate these differences, this project focuses on attitudes toward individual media companies. Specifically, previous research suggests that people who believe in harmful media effects will engage in preventive or accommodative strate- gies (Gunther et al., 2008), but the current research extends that argument by predict- ing that belief in harmful media effects will correlate with negative evaluations of media companies:

H2a: Belief in harmful third-person media effects from a media company will be negatively correlated with evaluations of that company. H2b: Belief in harmful overall media effects from a media company will be nega- tively correlated with evaluations of that company.

For this study, media companies will be evaluated through three constructs related to corporate image: CSR, word of mouth, and general satisfaction.

Although there are myriad definitions of CSR, the concept can generally be under- stood as business practices that provide benefits for groups and individuals other than the corporation itself. For example, Turker (2009) factor analyzed a set of CSR items and determined that CSR practices could be directed toward, for example, society at large, the natural environment, future generations, or employees. Considering the sur- vey design of this study, CSR will be assessed as it is directed toward customers (Turker, 2009). Presumably, people who believe that media companies produce media that are harmful to themselves and others will also believe that those companies are socially irresponsible, especially toward their customers.

The concept of word of mouth refers to the possibility that people will share infor- mation about a company with others. To the extent that the shared information is posi- tive, word of mouth is beneficial for corporations. In fact, Walsh and Beatty (2007) connected positive word of mouth with positive evaluations of service firms. Regarding belief in harmful media effects, it is unlikely that people will recommend companies

910 Journalism & Mass Communication Quarterly 93(4)

that they believe produce harmful media. Similarly, Walsh and Beatty (2007) found a strong positive relationship between customer satisfaction and aspects of corporate reputation. Satisfaction with media companies will likely be low for people who believe those companies are responsible for harmful media effects. Taken together, CSR, word of mouth, and satisfaction provide a varied picture of evaluations of media companies that is likely to be inversely related to belief in harmful media effects pro- duced by those media companies.

Method

To test the hypotheses stated above, an online survey was conducted through Amazon. com’s crowdsourcing service Mechanical Turk (MTurk). Recent research has shown that MTurk provides respondents for social science research who are more representa- tive of the population than many other forms of recruitment, including convenience sampling from university populations (e.g., Behrend, Sharek, Meade, & Wiebe, 2011; Berinsky, Huber, & Lenz, 2012; Buhrmester, Kwang, & Gosling, 2011). Respondents for the primary study (N = 375) were paid $0.75 for their participation in the survey; they were primarily male (61.9%) and ranged in age from 18 to 79 (M = 30.39, SD = 10.85). Some of the companies evaluated by respondents are primarily U.S.-based companies, so only respondents within the United States were allowed to complete the survey.

The survey began with demographic questions about age and gender. After these items, respondents were presented with items about three specific media companies from three different realms of media production: news media, film production, and video game production. The order in which the three media companies were presented was randomized to prevent order effects; questions about each media company were also randomly ordered.

Media Company Selection

Although the use of particular media companies may limit generalizability, it was nec- essary to specify media companies for respondents to evaluate those companies. Moreover, steps were taken to ensure that the media companies selected were familiar to respondents yet representative of similar media companies. Specifically, an Amazon MTurk pretest (n = 47), conducted roughly 2 weeks before the primary survey was used to determine which media companies to include in the primary survey. This pretest measured familiarity and liking of five major media companies from each of three con- sidered media categories: news media, film production, and video game production. For familiarity, each company within a particular category was ranked from 1 to 5; this ranking system was repeated for the other two categories of media company. Likeability was measured via a single seven-response semantic differential scale with dislike extremely and like extremely as the pole labels.1 Based on the results of the pretest sur- vey, the New York Times (NYT) was selected as the news media company, Universal Studios (US) as the film company, and Electronic Arts (EA) as the video game

Sherrick 911

company. These three companies were selected based on a high familiarity ranking (NYT: 3.09; US: 2.87; EA: 2.57) and a likeability score not significantly different than average, NYT: t(46) = 1.17, p = .246; US: t(46) = 1.17, p = .249; EA: t(46) = 1.24, p = .222. The highest ranking companies on the familiarity metric were not selected because they all showed likeability scores significantly different than the average, news: t(46) = 3.12, p = .003; film: t(46) = 5.97, p < .001; video game: t(46) = 4.72, p < .001.

Measures for the Primary Survey

Media effects. Perceptions of media effects were measured through two 5-point Likert- type items (1 = strongly disagree, 5 = strongly agree) for each of the three media companies. For each company, respondents were given the following statements: “Media products produced by [company name] are harmful to me” and “Media prod- ucts produced by [company name] are harmful to others” (see below for means and standard deviations). Scholars dispute the most appropriate way to analyze these self and other items in third-person effects research. Some (e.g., Cohen & Tsfati, 2009; Tsfati, Ribak, & Cohen, 2005) argue that self and other items should be included simultaneously as predictors of relevant outcomes, thereby controlling for self percep- tions when testing other perceptions (and vice versa). Others (e.g., McLeod et al., 1997; Neuwirth & Frederick, 2002; Schmierbach et al., 2011) argue the third-person perception is better operationalized as the difference between other and self measures, and overall belief in media effects should be operationalized as the sum of other and self variables.

After comparing four common approaches for analyzing third-person effects data, Schmierbach, Boyle, and McLeod (2008) recommended including both methods described above.2 As such, both methods are considered and reported below. To do so, the third-person perceptual gap was calculated as the difference between the other and self items, resulting in a perceptual gap variable for each company (NYT: M = 0.20, SD = 0.70; US: M = 0.17, SD = 0.76; EA: M = 0.17, SD = 0.78). And overall percep- tions of harmful effects were calculated by adding scores for both the self and other items for each company, resulting in an overall effects variable for each company (NYT: M = 4.36, SD = 1.71; US: M = 4.21, SD = 1.42; EA: M = 4.62, SD = 1.78).

CSR to customers. CSR was measured through the three items of the “CSR to Custom- ers” subscale of Turker’s (2009) CSR instrument. Turker (2009) developed the full scale to administer to corporate managers (who have more comprehensive knowledge of the corporations’ CSR practices), but the subscale used for the current study con- tains items specific to customer relations: “[Company name] provides full and accu- rate information about its products to its customers,” “[Company name] respects consumer rights beyond the legal requirements,” and “Customer satisfaction is highly important for [company name].” These items, as well as the other corporate image items, were submitted to a confirmatory factor analysis (see below), which supported the creation of a mean CSR scale (NYT: M = 3.38, SD = 0.77, Cronbach’s α = .81; US: M = 3.39, SD = 0.66, α = .74; EA: M = 2.84, SD = 1.01, α = .88). All corporate image

912 Journalism & Mass Communication Quarterly 93(4)

items, including the CSR items, were measured on 5-point Likert-type scales with strongly disagree and strongly agree as anchors.

Word of mouth. Word of mouth about the three companies was measured through three items as reported in Walsh and Beatty (2007). The items were “I’m likely to say good things about [company name],” “I would recommend [company name] to my friends and relatives,” and “If my friends were looking for a new company of this type, I would tell them to try [company name].” The Word-of-Mouth scale showed good reli- ability for each company (NYT: M = 3.31, SD = 1.02, α = .93; US: M = 3.51, SD = 0.75, α = .88; EA: M = 2.88, SD = 1.18, α = .96).

Satisfaction. General satisfaction with the three media companies was measured through three items as reported in Walsh and Beatty (2007). The items were “I am satisfied with the services [company name] provides,” “I am satisfied with my overall experience with [company name],” and “As a whole, I am not satisfied with [company name]” (reverse-coded). The Satisfaction scale showed acceptable reliability for each company (NYT: M = 3.47, SD = 0.88, α = .89; US: M = 3.77, SD = 0.63, α = .86; EA: M = 3.02, SD = 1.17, α = .95).

Results

Preliminary Analysis of Dependent Variables

Because the three scales used to measure corporate image were significantly corre- lated (see Table 1), the items of these scales were submitted to a confirmatory factor analysis, using maximum likelihood estimation in AMOS. When considering the three constructs as separate factors, the model for each company showed acceptable fit, NYT: χ2(24) = 40.12, p = .021, comparative fit index (CFI) = .994, root mean square error of approximation (RMSEA) = .042, 90% CI = [.017, .065], standardized root mean square residual (SRMR) = .02; US: χ2(24) = 51.05, p = .001, CFI = .986, RMSEA = .055, 90% CI = [.034, .076], SRMR = .03; EA: χ2(24) = 55.49, p < .001, CFI = .992, RMSEA = .059, 90% CI = [.039, .080], SRMR = .01. However, considering the high correlation between these factors, a unidimensional model was also considered, in which the nine items comprised a single scale. Analysis showed that the unidimen- sional model fit significantly worse on the data than the model with the three factors separated for all three companies, NYT, χdifference

2 ( )3 = 91.09, p < .001; US, χdifference 2 ( )3

= 139.12, p < .001; EA, χdifference 2 ( )3 = 120.65, p < .001. As such, the three scale mea-

sures of corporate image (CSR, Word of Mouth, and Satisfaction) are considered sepa- rately but simultaneously through multivariate regression.

Hypotheses Tests

Separate paired t tests were conducted to analyze the third-person perception for the three media companies (H1). NYT media products were seen as significantly more harmful to

Sherrick 913

others, M = 2.28, SD = 0.93, than to the self, M = 2.08, SD = .92, t(374) = −5.59, p < .001, Cohen’s d = 0.29; US media products were seen as significantly more harmful to others, M = 2.19, SD = 0.83, than to the self, M = 2.02, SD = .78, t(374) = −4.22, p < .001, d = 0.22; and EA media products were seen as significantly more harmful to others, M = 2.39, SD = 0.99, than to the self, M = 2.23, SD = .96, t(374) = −4.15, p < .001, d = 0.20. Thus, H1 regarding the third-person perceptual gap was consistently supported.

As noted above, there are two recommended methods for analyzing the effects of third-person effects variables: others and self variables simultaneously or perceptual gap and overall effects variables simultaneously. As such, two different multivariate

Table 1. Zero-Order Correlations Between Variables of Interest.

1 2 3 4 5 6

1. Harmful to me NYT US EA 2. Harmful to others NYT .71*** US .56*** EA .68*** 3. Others − self (TPE) NYT −.36*** .40*** US −.43*** .52*** EA −.38*** .43*** 4. Others + self (overall) NYT .92*** .93*** .03 US .88*** .89*** .07 EA .91*** .92*** .03 5. CSR NYT −.52*** −.49*** .03 −.54*** US −.21*** −.25*** −.06 −.26*** EA −.34*** −.38*** −.07 −.39*** 6. Word of mouth NYT −.50*** −.50*** −.01 −.54*** .78*** US −34*** −.36*** −.04 −.40*** .70*** EA −.39*** −.42*** −.05 −.44*** .85*** 7. Satisfaction NYT −.63*** −.61*** .01 −.67*** .80*** .83*** US −51*** −.49*** −.01 −.57*** .60*** .75*** EA −.43*** −.43*** −.01 −.47*** .85*** .91***

Note. NYT = New York Times; US = Universal Studios; EA = Electronic Arts; CSR = corporate social responsibility; TPE = third-person effects. *p < .05. **p < .01. ***p < .001.

914 Journalism & Mass Communication Quarterly 93(4)

regression models per media company were used to test the effect of belief in media effects on corporate image, with gender and age as control variables. However, as the two models only differ in that the same predictor variables (i.e., the self and other variables) are differently represented, the omnibus multivariate test for the two models is identical. This test was significant for all three companies, NYT: Wilks’s Λ = 0.53, F(12, 973.93) = 22.27, p < .001, ηp

2 = .19; US: Λ = 0.65, F(12, 973.93) = 14.16, p <

.001, ηp 2

= .13; EA: Λ = 0.70, F(12, 973.93) = 11.97, p < .001, ηp 2

= .11, so tests for both models are considered below.

The first model, which is used to test only H2a, included the unmodified self and other variables as predictors of the three corporate image variables: CSR, word of mouth, and satisfaction. According to analyses, there was a significant multivariate effect of both the self, NYT: Λ = 0.87, F(3, 368) = 18.57, p < .001, ηp

2 = .13; US: Λ = 0.88, F(3, 368) =

16.52, p < .001, ηp 2 = .12; EA: Λ = 0.94, F(3, 368) = 8.47, p < .001, ηp

2 = .07, and the

other, NYT: Λ = 0.92, F(3, 368) = 11.33, p < .001, ηp 2

= .09; US: Λ = 0.92, F(3, 368) = 11.14, p < .001, ηp

2 = .08; EA: Λ = 0.95, F(3, 368) = 7.19, p < .001, ηp

2 = .06, variables,

when considering all three dependent variables. Subsequent univariate tests (see Table 2) showed that perceptions of harmful effects on the self were negatively related to CSR (NYT: β = −.35, p < .001; US: β = −.10, p = .085; EA: β = −.17, p = .009), word of mouth (NYT: β = −.30, p < .001; US: β = −.20, p = .001; EA: β = −.22, p = .001), and satisfaction (NYT: β = −.40, p < .001; US: β = −.34, p < .001; EA: β = −.28, p < .001). Similarly, perceptions of harmful effects on others were negatively related to CSR (NYT: β = −.23, p < .001; US: β = −.19, p = .002; EA: β = −.28, p < .001), word of mouth (NYT: β = −.27, p < .001; US: β = −.24, p < .001; EA: β = −.27, p < .001), and satisfaction (NYT: β = −.32, p < .001; US: β = −.30, p < .001; EA: β = −.25, p < .001), lending support to H2a.

The second regression model, used to test both H2a and H2b, included the percep- tual gap variable (others − self) and the overall belief in media effects variable (others + self) as predictors of the three corporate image variables. For all three companies, the perceptual gap variable was not a significant multivariate predictor of the three dependent variables, NYT: Λ = 1.00, F(3, 368) = 0.65, p = .585, ηp

2 = .01; US: Λ =

0.99 F(3, 368) = 0.77, p = .514, ηp 2

= .01; EA: Λ = 0.99, F(3, 368) = 1.84, p = .139, ηp 2

= .02, suggesting H2a may not be supported. However, the overall media effects variable was a significant multivariate predictor of the dependent variables, NYT: Λ = 0.55, F(3, 368) = 99.05, p < .001, ηp

2 = .45; US: Λ = 0.67, F(3, 368) = 59.74, p < .001,

ηp 2

= .33; EA: Λ = 0.76, F(3, 368) = 38.81, p < .001, ηp 2

= .24. The univariate tests (see Table 3) showed that perceptions of overall media effects were negatively associ- ated with CSR (NYT: β = −.53, p < .001; US: β = −.26, p < .001; EA: β = −.40, p < .001), word of mouth (NYT: β = −.52, p < .001; US: β = −.39, p < .001; EA: β = −.45, p < .001), and satisfaction (NYT: β = −.66, p < .001; US: β = −.57, p < .001; EA: β = −.48, p < .001), showing consistent support for H2b.

Discussion

Before interpreting the effects of a belief in media effects on corporate image, it is important to discuss the differing results by analysis strategy for the belief in harmful

Sherrick 915

effects variables. Specifically, the first model reported above, which considers the harmful effects of media on self and others simultaneously (e.g., Cohen & Tsfati, 2009; Tsfati et al., 2005), suggests that belief in negative effects on others is negatively related to evaluations of the related media company. This model provides evidence consistent with H2a, predicting a negative relationship between belief in negative media effects for others and company evaluations, when also considering negative media effects for the self.

However, the second model, which considers the third-person perceptual gap and an overall belief in media effects consecutively (e.g., McLeod et al., 1997; Neuwirth & Frederick, 2002; Schmierbach et al., 2011), contradicts H2a as there was no relationship between the third-person perceptual gap and company evaluations, when also considering belief in overall media effects. In this model, it was instead a belief in the overall influence of media that was highly and negatively related to company evaluations. Importantly, the third-person perceptual gap was found in this study, so the people sampled do believe that others are more vulnerable to media effects than they are themselves. But this gap was not associated with the company evaluation variables. In other words, the corporate evalua- tions may not be driven by a desire to protect vulnerable others, but rather by the “broader” influence of presumed influence (Gunther & Storey, 2003)—in this case, the belief that harmful media effects are negative for everyone, including the self.

Table 2. Regression of Corporate Image Scales on Perceptions of First- and Third-Person Effects Variables (N = 375).

NYT US EA

B SE β B SE β B SE β

CSR Gendera 0.10 0.07 .06 0.07* 0.07 .05 0.30** 0.10 .15 Age −0.01* 0.00 −.10 −0.00 0.00 −.03 0.02*** 0.00 .17 Harmful to me −0.29*** 0.05 −.35 −0.09† 0.05 −.11 −0.15** 0.07 −.14 Harmful to others −0.20** 0.05 −.24 −0.15** 0.05 −.19 −0.30*** 0.07 −.29 Word of mouth Gendera 0.22* 0.09 .10 0.08 0.08 .05 0.28* 0.11 .12 Age −0.01† 0.00 −.08 −0.01† 0.00 −.10 0.01† 0.01 .09 Harmful to me −0.33*** 0.07 −.30 −0.19** 0.06 −.20 −0.25*** 0.08 −.20 Harmful to others −0.31*** 0.07 −.29 −0.23*** 0.05 −.25 −0.34*** 0.08 −.28 Satisfaction Gendera 0.19** 0.07 .11 0.00 0.06 .00 0.32** 0.11 .13 Age −0.01 0.00 −.06 −0.00 0.00 −.03 0.02** 0.01 .14 Harmful to me −0.38*** 0.05 −.40 −0.28*** 0.04 −.35 −0.32*** 0.08 −.26 Harmful to others −0.31*** 0.05 −.33 −0.23*** 0.04 −.30 −0.30*** 0.07 −.25

Note. NYT = New York Times; US = Universal Studios; EA = Electronic Arts; CSR = corporate social responsibility. aFemale = 1. †p < .10. *p < .05. **p < .01. ***p < .001.

916 Journalism & Mass Communication Quarterly 93(4)

Although the third-person effects and influence of presumed media influence hypoth- eses are meant to be complementary—rather than competing—predictions (Gunther & Storey, 2003), these data show that the effects of these two beliefs are not necessarily equivalent. In this research, only the influence of presumed (harmful) media influence translates into negative evaluations of the companies that create that media. Specifically, individual media companies were negatively evaluated by people who believed that those companies’ media products were harmful to the self and others. This finding extends the existing research on the effects of a belief in media effects. Prior research has typically shown that a belief in (negative) media effects is correlated with a general will- ingness to censor media or media types (i.e., pornography or video games; for example, McLeod et al., 1997; Salwen, 1998; Schmierbach et al., 2011). The current study pro- vides the more specific expectation that individual media companies may be targets of preventive or accommodative strategies caused by a belief in media effects. It also iden- tifies corporate image variables as a potentially important outcome of a belief in media effects. Future research on third-person effects and the influence of presumed media influence should be cognizant of the outcome of study and the target of that outcome, especially considering the differing results of the analyses performed here.

The fact that belief in overall harmful effects, rather than third-person effects, is negatively related to corporate evaluations may actually be worse news for media

Table 3. Regression of Corporate Image Scales on Perceptions of Perceptual Gap and Overall Effects Variables (N = 375).

NYT US EA

B SE β B SE β B SE β

CSR Gendera 0.10 0.07 .06 0.07 0.07 .05 0.31** 0.10 .15 Age −0.01* 0.00 −.10 −0.00 0.00 −.03 0.02*** 0.00 .17 Perceptual gap 0.05 0.06 .05 −0.03 0.04 −.03 −0.07 0.06 −.05 Overall effects −0.24*** 0.02 −.53 −0.12*** 0.02 −.26 −0.22*** 0.03 −.39 Word of mouth Gendera 0.22* 0.09 .10 0.08 0.08 .05 0.28* 0.11 .12 Age −0.01† 0.00 −.08 −0.01* 0.00 −.10 0.01† 0.01 .09 Perceptual gap 0.01 0.06 .01 −0.02 0.05 −.02 −0.05 0.07 −.03 Overall effects −0.32*** 0.03 −.54 −0.21*** 0.03 −.40 −0.29*** 0.03 −.44 Satisfaction Gendera 0.19** 0.07 .11 0.00 0.06 .00 0.32** 0.11 .13 Age −0.01 0.00 −.06 −0.00 0.00 −.03 0.02** 0.01 .14 Perceptual gap 0.04 0.05 .03 0.02 0.04 .02 0.01 0.07 .01 Overall effects −0.35*** 0.02 −.68 −0.25*** 0.02 −.56 −0.31*** 0.03 −.47

Note. NYT = New York Times; US = Universal Studios; EA = Electronic Arts; CSR = corporate social responsibility. aFemale = 1. †p < .10. *p < .05. **p < .01. ***p < .001.

Sherrick 917

companies. If the negative evaluations were simply associated with a desire to protect vulnerable others (e.g., children), then media companies could specifically counteract that phenomenon by emphasizing their efforts to protect vulnerable populations—for example, through compliance with content rating systems. However, as the negative evaluations are connected to a general belief in harmful effects of media, media com- panies will need to take more pervasive steps to counteract or alleviate the negative impact of belief in negative media effects.

One important first step should be to become directly involved in debates about media effects; this research shows that media companies have a vested interest in the outcomes of these debates. Specifically, if these debates about media effects encour- age belief in harmful media effects—as they often do, this could lead to negative per- ceptions of CSR, less positive word of mouth, and general dissatisfaction, all outcomes that could negatively affect the companies’ bottom line. Public relations departments and other visible entities of media companies should engage in the debate over media effects with these outcomes in mind.

Perceptions of harmful media effects are likely to be associated with social irre- sponsibility, but media companies could emphasize the positive outcomes of their media as evidence of their societal value. Or, if a media company is believed to have caused harmful effects, it can show social responsibility by counteracting that effect through, for example, counseling for those affected or education for those who may be vulnerable.

To ensure that word of mouth is more positive than negative, media companies should directly engage vocal customers and opponents. This may require debating the strong voices that consistently blame media for negative real-world consequences. Although, there is some evidence of serious negative media effects, media companies need to, again, emphasize the potential positive effects of their products in these debates. Word of mouth is likely to be positive only insomuch as those talking have positive things to talk about.

In this study, satisfaction with media companies was shown to negatively correlate with perceptions of harmful media effects; however, this outcome may also be associ- ated with the overall quality of the company’s media products. Future research, includ- ing that done by media companies, should investigate how the general quality of media products is related to perceptions about the harm or benefit of those products. It is possible that the two evaluations are related, that is, that people think media products that they do not like are also harmful. If so, media companies should focus on divorc- ing those two concepts.

Media companies could also become more directly involved in the debate over media effects by funding relevant research. The general belief in harmful media effects is likely, at least in part, motivated by horror stories from academic research examin- ing, for example, the link between violent media and aggression. However, this vocal camp of media effects researchers may overshadow a significant portion of scholars who are interested in the benefits of media use. For example, particular forms of news media have been connected to decreased stigmatization of out-groups (Oliver, Dillard, Bae, & Tamul, 2012), films have been shown to provide greater insight and meaning

918 Journalism & Mass Communication Quarterly 93(4)

about life (Oliver & Raney, 2011), and use of prosocial video games has been con- nected to subsequent prosocial behaviors (Gentile et al., 2009). If this research into the benefits of media becomes more popular in academia and pervasive in the societal conversation, it may counteract the thus-far dominant argument about the negative effects of media. Media companies should take an active role in ensuring this happens.

Thus far, the conversations over media effects have largely been dominated by poli- ticians and academics that may or may not have the media industries’ best interests in mind. This can lead to some important misperceptions about the media industries. The video game industry, for example, is often seen as promoting violence and other base desires, despite the increasing diversity of game products (e.g., Evangelista, 2013). This perception, as well as other negative views of media, is unlikely to change unless the media industries make a concerted effort to counteract those perceptions. Moreover, the current research suggests that individual companies—not just the industry writ large—should become actively involved in this discussion as a belief in harmful media effects can lead to negative company-specific evaluations.

Limitations

As with any correlational data, the suggested ordering of the relevant variables cannot be empirically proven. In other words, it is possible that negative evaluations of media companies lead to the belief that those companies produce harmful media, although this alternative is less logically attractive than the opposite. Moreover, a wealth of research on the effects of media effects suggests that belief in harmful media effects leads to preventive or accommodative strategies (Perloff, 2009), such as negative eval- uations of media companies. Perhaps most relevant is Tal-Or et al.’s (2010) experi- mental study establishing the causal direction of this relationship. That said, the novelty of the outcomes examined here precludes a definitive causal statement regard- ing which variable comes first.

This study uses only one possibility of the numerous conceptualizations of corpo- rate image. Although CSR, word of mouth, and satisfaction are important variables that can provide a varied understanding of perceptions of corporate image, there are likely aspects of corporate image (e.g., economic viability, loyalty) that are ignored through this particular conceptualization. This study is not meant to provide a final word on the connection between belief in media effects and corporate image, but it does provide an initial understanding of this relationship.

Although the argument presented here claims that preventive or accommodative strategies in response to belief in media effects can be varied, only one potential outcome is considered (evaluations of media companies). Negative evaluations may also be a mild outcome, when compared with other studied outcomes, like direct calls for censorship. However, for media companies, consumer evaluations are nontrivial, as they are likely to correspond directly with behaviors like purchase intention (e.g., Glasman & Albarracín, 2006; Webb & Sheeran, 2006). Moreover, as Gunther et al. (2008) pointed out, many outcomes of presumed media influence are

Sherrick 919

“often described under the umbrella of the so-called behavioral component,” so future research should continue to disentangle the myriad possible attitudinal and behavioral outcomes of a belief in media effects. Comparing these different out- comes in a single study could be worthwhile, particularly in determining what groups (e.g., media companies, governments) are especially targeted with preven- tive or accommodative strategies.

Finally, the use of MTurk for sampling respondents may limit the generalizability of this research. Previous research has shown that MTurk samples are demographi- cally diverse and similar or superior to other commonly used samples (e.g., Behrend et al., 2011; Berinsky et al., 2012; Buhrmester et al., 2011), but sufficient demographic data were not collected to be able to make similar claims about the particular respon- dents for this study.

Conclusion

The results of these survey data show that people who believe that media companies are responsible for harmful media effects are likely to evaluate those media companies more negatively. While this may seem an intuitive conclusion, the data here provide empirical support for this assumption and should encourage media companies to more carefully consider their roles in the media effects debate. This research also includes important lessons for researchers in the third-person effects domain. Namely, these data show that the third-person effects perceptual gap and the influence of presumed media influence do not necessarily result in identical outcomes, which should encour- age caution for researchers making claims about these two theories. Moreover, this research extends the research on the effects of media effects by testing a new potential outcome of belief in media effects: negative evaluations of media companies.

Author’s Note

This research was previously presented at the 2013 annual conference for the Association for Education in Journalism and Mass Communication in Washington, D.C.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Arthur W. Page Center for Integrity in Public Communication.

Notes

1. Due to a computer error during data collection, only four film and four video game compa- nies had likeability data and were used for likeability comparisons.

920 Journalism & Mass Communication Quarterly 93(4)

2. Sun et al. (2008) provided a modified version—also including so-called first-person per- ceptual gap (self–others)—of the method suggested by Schmierbach, Boyle, and McLeod (2008). Considering the focus of this project is on harmful effects, which are likely to relate to the traditional third-person perceptual gap, Sun, Shen, & Pan’s (2008) modification is not considered in these analyses.

References

Behrend, T. S., Sharek, D. J., Meade, A. W., & Wiebe, E. N. (2011). The viability of crowdsourc- ing for survey research. Behavioral Research, 43, 800-813. doi:10.3758/s13428-011-0081-0

Berinsky, A. J., Huber, G. A., & Lenz, G. S. (2012). Evaluating online labor markets for experimental research: Amazon.com’s Mechanical Turk. Political Analysis, 20, 351-368. doi:10.1093/pan/mpr057

Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3-5. doi:10.1177/1745691610393980

Cohen, J., & Tsfati, Y. (2009). The influence of presumed media influence on strategic voting. Communication Research, 36, 359-378. doi:10.1177/0093650209333026

Cohen, J., Tsfati, Y., & Sheafer, T. (2008). The influence of presumed media influence in poli- tics: Do politicians’ perceptions of media power matter? Public Opinion Quarterly, 72, 331-344. doi:10.1093/poq/nfn014

Cohen, J., & Weimann, G. (2008). Who’s afraid of reality shows? Exploring the effects of per- ceived influence of reality shows and the concern over their social effects on willingness to censor. Communication Research, 35, 382-397. doi:10.1177/0093650208315964

Evangelista, B. (2013, January 27). Video games drawn into violence debate. San Francisco Chronicle. Retrieved from http://www.sfgate.com/news/article/Video-games-drawn-into- violence-debate-4219013.php#page-2

Gentile, D. A., Anderson, C. A., Yukawa, S., Ihori, N., Saleem, M., Ming, L. M., . . . Sakamoto, A. (2009). The effects of prosocial video games on prosocial behaviors: International evi- dence from correlational, longitudinal, and experimental studies. Personality and Social Psychology Bulletin, 35, 752-763. doi:10.1177/0146167209333045

Glasman, L. R., & Albarracín, D. (2006). Forming attitudes that predict future behavior: A meta-analysis of the attitude-behavior relation. Psychological Bulletin, 132, 778-822. doi:10.1037/0033-2909.132.5.778

Gunther, A. C., Bolt, D., Borzekowski, D. L. G., Liebhart, J. L., & Dillard, J. P. (2006). Presumed influence on peer norms: How mass media indirectly affect adolescent smoking. Journal of Communication, 56, 52-68. doi:10.1111/j.1460-2466.2006.00002.x

Gunther, A. C., Perloff, R., & Tsfati, Y. (2008). Public opinion and the third-person effect. In W. Donsbach & M. W. Traugott (Eds.), The SAGE handbook of public opinion research (pp. 184-192). London, England: SAGE. doi:10.4135/9781848607910.n18

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

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

Neuwirth, K., & Frederick, E. (2002). Extending the framework of third-, first-, and sec- ond-person effects. Mass Communication and Society, 5, 113-140. doi:10.1207/ S15327825MCS0502_2

Sherrick 921

Oliver, M. B., Dillard, J. P., Bae, K., & Tamul, D. J. (2012). The effect of narrative news format on empathy for stigmatized groups. Journalism & Mass Communication Quarterly, 89, 205-224. doi:10.1177/1077699012439020

Oliver, M. B., & Raney, A. A. (2011). Entertainment as pleasurable and meaningful: Identifying hedonic and eudaimonic motivations for entertainment consumption. Journal of Communication, 61, 984-1004. doi:10.1111/j.1460-2466.2011.01585.x

Park, S.-Y. (2005). The influence of presumed media influence on women’s desire to be thin. Communication Research, 32, 594-614. doi:10.1177/0093650205279350

Paul, B., Salwen, M. B., & Dupagne, M. (2000). The third-person effect: A meta-analysis of the perceptual hypothesis. Mass Communication and Society, 3, 57-85. doi:10.1207/ S15327825MCS0301_04

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 (pp. 252-268). New York, NY: Routledge.

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, 163-186.

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

Scharrer, E., & Leone, R. (2006). I know you are but what am I? Young people’s perceptions of varying types of video game influence. Mass Communication and Society, 9, 261-286. doi:10.1207/s15327825mcs0903_1

Scharrer, E., & Leone, R. (2008). First-person shooters and the third-person effect. Human Communication Research, 34, 210-233. doi:10.1111/j.1468-2958.2008.00319.x

Schmierbach, M., Boyle, M. P., & McLeod, D. M. (2008). Understanding person percep- tions: Comparing four common statistical approaches to third-person research. Mass Communication and Society, 11, 492-513. doi:10.1080/15205430802375311

Schmierbach, M., Boyle, M. P., Xu, Q., & McLeod, D. M. (2011). Exploring third-person differences between gamers and nongamers. Journal of Communication, 61, 307-327. doi:10.1111/j.1460-2466.2011.01541.x

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

Sun, Y., Shen, L., & Pan, Z. (2008). On the behavioral component of the third-person effect. Communication Research, 35, 257-278. doi:10.1177/0093650207313167

Tal-Or, N., Cohen, J., Tsfati, Y., & Gunther, A. C. (2010). Testing causal direction in the influence of presumed media influence. Communication Research, 37, 801-824. doi:10.1177/0093650210362684

Tsfati, Y., & Cohen, J. (2005). The influence of presumed media influence on demo- cratic legitimacy: The case of Gaza settlers. Communication Research, 32, 794-821. doi:10.1177/0093650205281057

Tsfati, Y., Ribak, R., & Cohen, J. (2005). Rebelde Way in Israel: Parental perceptions of television influence and monitoring of children’s social and media activities. Mass Communication and Society, 8, 3-22. doi:10.1207/s15327825mcs0801_2

Turker, D. (2009). Measuring corporate social responsibility: A scale development study. Journal of Business Ethics, 85, 411-427. doi:10.1007/s10551-008-9780-6

Walsh, G., & Beatty, S. E. (2007). Customer-based corporate reputation of a service firm: Scale development and validation. Journal of the Academy of Marketing Science, 35, 127-143. doi:10.1007/s11747-007-0015-7

922 Journalism & Mass Communication Quarterly 93(4)

Webb, T. L., & Sheeran, P. (2006). Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychological Bulletin, 132, 249- 268. doi:10.1037/0033-2909.132.2.249

Wolf, H., & Dee, J. (2013). Public opinion regarding the role of government in regulating violent content in video games. First Amendment Studies, 47(2), 170-197. doi: 10.1080/ 21689725.2013.852795

Author Biography

Brett Sherrick (PhD, Penn State) is an instructor in the Department of Journalism at the University of Alabama. His research interests include video games, persuasion, and media effects. He is primarily interested in investigating why people play games and how they benefit from those gameplay experiences.