Issue Importance,Third-Person Effects of Protest News,and Participation in Taiwan’s Sunflower Movement

Abstract

How do news coverage of a grassroot protest movement and perceived importance of the movement affect people’s participation in the movement? And does people’s inference of the effect of the news on themselves versus others make a difference in participation? Informed by the third-person effect hypothesis, we examine these questions in the context of the student-led Sunflower Movement in Taiwan that rose in opposition to a trade pact with China. In the study, we advanced three propositions: First, that the perceived effects of the protest news on oneself would be a better predictor of political participation than would perceived effects of such news on others. Second, that the perceived effect on oneself, not on others, would enhance the impact of issue importance on participation in the movement. And third, how people processed protest news would be another intermediate mechanism on subsequent participatory activities. We found support for these propositions in data collected from a probability sample of 1,137 respondents. The implications of the findings for the robust third-person effect research are discussed.

Keywords

third-person effect issue importance political participation news attention news elaboration

Introduction

How do news coverage of a grassroot protest movement and perceived importance of the movement affect people’s participation in the movement? Furthermore, does people’s inference of the effect of the news on themselves versus others make a difference in participation? These are some of the enduring concerns of media effect research in general and the third-person effect hypothesis in particular. In the era of “horizontal media” (e.g., social media; see Shaw, Hamm, & Terry, 2006), which provide citizens unprecedented opportunities to participate in social activism and politics, these concerns call for more robust research.

Informed by the third-person effect hypothesis and the cognitive mediation model of news processing, we examine these research concerns in the context of the student-led Sunflower Movement in Taiwan that opposes a trade pact with China. Specifically, we propose the more that respondents pay attention to the news and think about its ramifications, the more they will believe themselves to be impacted, which leads to differential perceptual effects on oneself and on others. Furthermore, we propose that the perceived effect on oneself, not on others, will enhance the impact of issue importance on participation in the movement. Because the third-person effect literature reported inconsistent results regarding how the perceived effects on the self and on others trigger behavior responses (Golan & Day, 2008; Lo & Wei, 2002; Rojas, 2010; Wei, Lo, & Lu, 2010), clarifying the differential role of perceived effects on the self and perceived effects on others will contribute to the literature.

More importantly, although the third-person effect has been a widely tested theory with ample empirical evidence supporting it (Bryant & Miron, 2004; Perloff, 2009), few studies have attempted to assess the influence of news media on oneself relative to others from an integrated approach that incorporates issue importance and news processing variables. This study aims to fill the gap by developing a more comprehensive model that articulates the theoretical linkages among perceived issue importance, news attention, news elaboration, the third-person effect, and political participation.

Furthermore, the growing research on the behavioral component of the third-person effect has primarily focused on attitudes and behavioral intentions. Very few studies have assessed the impact of perceived effects on actual behavior (Cohen & Tsfati, 2009; Rojas, 2010). By incorporating actual behavior such as online discussions of and participation in the protests into the third-person effect framework, the findings of the study contribute to the third-person effect literature because they expand the scope of the behavioral component of the third-person effect hypothesis.

Finally, by constructing a model that reflects the theoretical linkages among perceived issue importance, information-processing variables, and perceived effects on self and actual behavior, this study makes a contribution to the understanding of both the perceptual and behavioral components of the third-person effect hypothesis. Based on an expanded cognitive mediation model, which suggests that motivation, attention, and elaboration are the information-processing variables that determine learning from the news (Beaudoin & Thorson, 2004; Eveland, 2001, 2002), we propose that levels of perceived issue importance, news attention, and news elaboration of the protest news will affect individuals’ judgment of the effects of the news on the self relative to others. The perceived effects on oneself will in turn affect participation in the movement, online and offline.

Context of Study

The Sunflower Movement erupted in Taiwan in spring 2014. College students, added by civic groups, launched a series of protests against the proposed trade deals with mainland China. Although advocates of the trade pact argued that the pact would benefit Taiwan’s economy by allowing the island and mainland China greater access to each other’s service sectors, the protesters claimed that the trade pact would amount to “betraying Taiwan” by letting the mainland exert its economic and political influence (Fung, 2014). They also argued that the trade pact would benefit large companies at the expense of small- and medium-sized local businesses (Chang, 2014). College students feared that pact would mean high unemployment, which would rob them of the earning power to buy pricey homes (Lin, 2015). Protesters occupied Taiwan’s Legislature to disrupt the passing of the Cross-Strait Service Trade Agreement. At the height of the movement on March 30, protesters claimed that nearly half a million people participated in the protest.

Because the movement marked the first time that the Legislature building had been occupied by college students, the movement received wide coverage in Taiwan’s news media. Due to the extensive media exposure, two of the student leaders, Lin Fei-Fan and Chen Wei-Ting, became highly recognized public figures overnight. The movement thus provided a fitting context to test the propositions regarding the effects of the protest news on participation that works through cognitive mechanisms and perception.

Literature Review and Hypotheses

The Third-Person Effect

The third-person effect hypothesis, proposed by Davison (1983), states that people have the propensity to believe that others are more likely to be impacted by media messages than themselves. Davison’s third-person effect hypothesis has led to numerous studies that have supported the perceptual component of the hypothesis using both surveys and experiments (Andsager & White, 2007; Perloff, 1993, 1999, 2009).

Recent studies that extended media messages from entertainment (e.g., violent media and pornography) to news reporting of current events also documented strong support for the perceptual component of the hypothesis. Focusing on news coverage of the bird flu, for example, Wei, Lo, and Lu (2007) reported that individuals perceived a greater effect of the news on other people than themselves. Similar results were reported in a tainted food scandal in Taiwan. Wei et al. (2010) found that respondents judged others as more influenced by the news coverage of the recalls than they were. Consistent results were reported in other studies that found respondents perceived the effect of news coverage to be greater on other voters than on themselves (Price & Stroud, 2005).

In the context of news about the student-led protests in Taiwan, we expected that respondents would perceive such news to have a greater effect on others than on themselves.

H1: Respondents will perceive news about the student protests of the China trade pact to have a greater effect on others than on themselves.

Perceived Issue Importance and Third-Person Effect

The cognitive mediation model, proposed by Eveland (2001, 2002), states that motivation, attention, and elaboration are the key information-processing factors that affect learning from the news. When people are motivated to process information about an event, they should be more attentive to the media message and more likely to engage in elaborative processing and thus should acquire more information from media.

According to Eveland (2001, 2002), motivation functions as an antecedent of information processing. Past studies found that motivation, operationalized as surveillance gratification sought from news media, was significantly associated with attention and elaboration (Eveland, 2001; Eveland, Shah, & Kwak, 2003). However, surveillance gratification sought from news media influences media effects only indirectly through information processing (Eveland, 2001; Eveland et al., 2003). Eveland (2001) acknowledged the need to expand the cognitive mediation model by including other dimensions of motivations beyond surveillance gratification sought from news media.

Because past research has demonstrated that issue importance motivates individuals to engage in information-processing behaviors (Eagly & Chaiken, 1993; Griffin, Dunwoody, & Neuwirth, 1999), we seek in our study to expand the cognitive mediation model by incorporating perceived issue importance as a motivation for processing the protest news. Petty and Cacioppo (1979) found that high issue involvement enhances subsequent information processing and thinking about the messages. A recent study by Lo, Wei, Lu, and Hou (2015) reported that the level of perceived issue importance of controversial importation of beef affected third-person perception. That is, the more the controversy was perceived as important, the more respondents believed themselves affected by the news about it.

Along the lines of the cognitive mediation model and third-person effect hypothesis, we expect that perceived issue importance of the controversial China trade pact will have a positive impact on news attention, news elaboration, and perceived effects of the protest news. When perceived issue importance is high, it is more likely that people will pay attention to news about the student protests, elaborate on the ramifications, and infer the effects of such news on themselves and others. On the contrary, when perceived issue importance is low, it is less likely that people will engage in systematic processing of the protest news and believe such news to affect themselves and others.

These theoretical expectations are consistent with the Elaboration Likelihood Model of Persuasion framework. As Petty and Cacioppo (1986) proposed, persuasion may occur via either “central” or “peripheral” routes, depending on the personal relevancy of media messages. They elaborated that when personal relevancy of media messages is high, the central route is activated and the individual becomes highly motivated to scrutinize and evaluate the implications of the information. However, when personal relevancy of media messages is low, the peripheral route is engaged and the individual is not motivated to process the information thoroughly. Based on the cognitive mediation model and the elaboration likelihood model of persuasion, we propose that when the importance of an issue is perceived as high, individuals will be more likely to actively seek media messages and process them thoroughly, which will result in greater perceived effects of the message. That is, when individuals are informed by the news media about something important to think about, they will be more likely to give thoughts to them and internalize the news. On the contrary, when perceived issue importance is low, individuals will not be motivated to process media messages deeply and will be less likely to acknowledge that the issue influences them or others.

Past research has consistently shown that issue importance affects the intensity of information seeking and processing (Eagly & Chaiken, 1993; Griffin et al., 1999; Petty, Brinol, & Priester, 2009). When issue importance or personal relevance of the message is high, people will be more likely to pay attention to and engage in greater elaboration of the message and thus will perceive a greater impact of the message on oneself and others (Petty et al., 2009). Accordingly, we hypothesize the following:

H2a: Perceived issue importance will be significantly associated with attention to news about the student protests of the China trade pact.

H2b: Perceived issue importance will be significantly associated with elaboration of news about the student protests of the China trade pact.

H2c: Perceived issue importance will be significantly associated with perceived effects of news about the student protests on oneself and others.

Cognitive Mechanisms and the Third-Person Effect

We also anticipate that news attention and news elaboration will affect the third-person effect, as previous research has found that how people process news affects their appraisal of the effects of news on others and themselves (Neuwirth, Frederick, & Mayo, 2002). Following the logic of the cognitive mediation model (Eveland, 2001, 2002) as well as the elaboration likelihood model of persuasion (Petty & Cacioppo, 1986), we propose that when people are motivated to process information that they perceived as important, they are more likely to devote a large amount of attention to the news about the issue, engage in elaborative processing of the news, and then comprehend the news.

Following this logic, we expect that both news attention and news elaboration will lead to greater perceived effects of the news on oneself and others. Therefore, the more deeply people process the news in terms of paying more attention and thinking about its implications, the more they perceive themselves and other people as being affected by the news.

In a study of news about the bird flu outbreak, Wei et al. (2007) demonstrated that news attention and elaboration were positively related to both perceived effects of the bird flu news on oneself and on others. Similar results were reported in a more recent study of perceived effects of news concerning the safety of imported U.S. beef in Taiwan. Lo et al. (2015) also reported that attention to news about the beef controversy was positively related to perceived effects of such news on oneself. In addition, news elaboration, which refers to one’s inclination to think about a message (Petty & Cacioppo, 1986), was positively related to perceived effects of such news on oneself and on others. By relating incoming information to one’s existing knowledge, elaboration was found to result in a greater perceived impact of media messages on oneself and others (Wei et al., 2010).

In the context of news about the student protests against the China trade pact, the coverage was extensive. We estimate that the salience of the protest news will lead to perceived issue importance. Therefore, we expect individuals who consider the issue as important will be more likely to pay attention to the news and think about its personal ramifications. Consequently, they tend to perceive a greater impact of the news on oneself and others. In light of these considerations, we hypothesized the following:

H3: Attention to news about the student protests will be positively related to perceived effects of such news on oneself and others.

H4: Elaboration of news about the student protests will be positively related to perceived effects of such news on oneself and others.

Perceived Media Effects and Political Participation

Participation in politics includes a broad range of citizen activities such as expressing opinion, discussions, and protesting or voting. Conge (1988) differentiated mobilized and voluntary participation. The former refers to behavior sponsored by the government (e.g., voting), whereas the latter refers to action initiated by citizens in pursuit of their own interest (e.g., protesting). Regardless of type, participation in the political sphere aims to influence public policy.

To address how third-person effects and information-process variables affect the perceived effects and participation nexus, we propose that perceived effects on oneself will be a better predictor of political behavior than will perceived effects on others. This expectation is derived from the proposition advanced in the extended parallel process model (Witte, 1994) and evidence in previous third-person effect research. According to the extended parallel process model, if a threat is perceived as personally relevant or important, people will be motivated to take protective or preventive actions (Witte, 1994; Witte & Morrison, 2000). In the context of the student-led protests against the China trade pact, news coverage of the protests was intensive, including the objection to the clause in the pact which might cause loss of jobs to the mainland. College students in Taiwan feared such a prospect, which would impact them negatively. The intensity of news coverage thus would alert college students of the protests as personally important and would motivate them to take political action. The more that students perceived the news about the protests to affect themselves, the likelier that they would participate in the protests.

Recent studies (e.g., Liu & Lo, 2014; Lo et al., 2015; Wei et al., 2010) on the third-person effect of news messages documented a consistent pattern that it is the perceived effects on oneself that prompt individuals to take action out of self-interest in assessing the impact of media messages that are personally relevant. Studies about food recall news, swine flu news, and an imported beef controversy documented that perceived effects on oneself resulted in information seeking and taking protective actions (Liu & Lo, 2014; Lo et al., 2015; Wei et al., 2010). Wei and Lo (2015) concluded that citizen voluntary action based on self-interest, and aimed at influencing the government’s decision making, is consistent with the third-person effect theory in explaining people’s political behavior. In the words of Golan and Day (2008), certain messages do not affect others enough. Therefore, perceived effects on the self would be a better predictor of political behavior than would perceived effects on others because politically involved individuals may be willing to act only if they perceive themselves to be affected by political messages. Accordingly, we proposed that perceived effects on oneself, as compared with perceived effects on others, are more likely to result in participation in the protest movement.

H5a: Perceived effects of the protest news on oneself will be more strongly and positively related to discussions of the protests than will perceived effects on others.

H5b: Perceived effects of the protest news on oneself will be more strongly and positively related to participation in the protests than will perceived effects on others.

The Theoretical Model

This study integrates the cognitive mediation model with the third-person effect hypothesis by building a new model that articulates the theoretical linkages among issue importance, key information-processing variables, perceived effects of the protest news, and political participation. We are particularly interested in exploring whether and how news information-processing variables and perceived effects of the protest news enhance or mitigate the impact of perceived issue importance on political behavior.

As Figure 1 shows, perceived issue importance of the student protests over the China trade pact provides stimuli for cognitive processing (i.e., attention to and elaboration) of the protest news; this leads to perceived effects on oneself, which will motivate respondents to engage in political discussions and take political actions such as mobilizing support and participating in the protests.

Figure 1.

The proposed theoretical model.

Based on the model, we raised RQ1:

RQ1: Will news information-processing variables and perceived effects of the protest news on oneself and others enhance or reduce the impact of issue importance on political participation?

Method

Sampling

A large-scale survey with a probability sample of college students was conducted to collect data in April 2014 when the movement had just ended. Because the Sunflower Movement was characterized as a series of student-led protests, we deemed it appropriate to target college students on the island as our population of study. To draw the sample, a multistage cluster sampling strategy was used, in which 12 colleges, both public and private, were randomly selected from a pool of 147 colleges across Taiwan in the first stage. In the next stage, considering general educational classes are open to all majors, three general education classes were randomly chosen in each selected college. A master list of course offerings for the spring semester of 2014 was consulted in this step. Finally, with prior permission of instructors, self-administered questionnaires were provided in the selected classes. The surveys took place in a 3-week period starting on April 16 and ending on May 7. Participation was voluntary, no course credit was offered, and respondents were assured of confidentiality and anonymity. Trained graduate students under the supervision of one of the authors assisted in data collection. A total of 1,200 questionnaires were distributed, from which 1,137 (94.75%) completed ones were returned.

Of the sample, gender was evenly distributed. A total of 548 (48.4%) of the respondents were male and 585 were female (51.6%). The average age of the sample was 20.46 (SD = 1.34, ranging from 18 to 30). In terms of class standing, roughly a fifth (17.1%) were freshmen, a third (33.5%) sophomores, another third (32.4%) juniors, and the remaining fifth (17.1%) were seniors.¹

Measurement of Key Variables

Perceived effects of news about the student protests on oneself and others

Respondents were first asked to indicate to what extent news about student protests of the China trade pact influenced them in terms of (1) attitudes toward the student-led protests, (2) attitudes toward the government’s handling of the student-led protests, (3) attitudes toward the police’s handling of the student-led protests, and (4) attitudes toward the China trade pact. The measure of the perceived effect of news about the student protests on others consisted of four parallel items (replacing “you” and “yourself” with “others” and “themselves”). The response categories ranged on a 5-point scale from 1 (“no influence at all”) to 5 (“a great deal of influence”). Results of a principal components factor analysis showed that the self and other items were clearly grouped into two distinct factors. The two-factor solution accounted for 77.29% of the total variance. The four “others” items, which loaded on the first factor (Eigenvalue = 3.21; accounting for 40.11% of the variance), were averaged to create a measure of “perceived effects on others” (M = 3.61, SD = 0.86, α = .92). The second factor, which contained the four “self” items (accounting for 37.72% of the variance; Eigenvalue = 3.02), was also averaged to form an index of “perceived effect on oneself” (M = 3.20, SD = 1.00, α = .89).

Perceived issue importance

To assess the perceived importance of student protests of the China trade pact, on a 5-point Likert-type scale (1 = “strongly disagree” and 5 = “strongly agree”), respondents were asked to indicate their agreement with these statements: (1) “I would think that the student-led protest over the China trade pact is an important political issue”; (2) “I would think that the student-led protest over the China trade pact is an important economic issue”; and (3) “I would think that the student-led protest over the China trade pact is an important social issue.” Principal components analysis showed that the three items were grouped in a single factor, thus indicating that they measured the same underlying concept. The single-factor solution explained 65.17% of the total variance (Eigenvalue = 2.61). The average of the three items was used to create a composite index of “perceived issue importance” (M = 3.98, SD = 0.74, α = .74).

Attention to student protest news

Attention to student protest news was measured with three items adopted from previous studies (Wei et al., 2010). Respondents were asked to indicate how much attention they paid to news about the student-led protests on television, in newspapers, and on the Internet. The 5-point response categories ranged from 1 (“no attention at all”) to 5 (“a great deal of attention”). Results of a principal components analysis confirmed that the three items were loaded in a single factor. The one-factor solution explained 62.78% of the variance (Eigenvalue = 2.51). A composite measure of “attention to student protest news” was constructed by adding the three items and dividing by three (M = 3.61, SD = 0.90, α = .76).

Elaboration of the student protest news

To measure elaboration, respondents were asked to indicate their agreement with four statements adopted from previous research (Eveland, 2001, 2002) on a 5-point Likert-type scale, where 1 meant “strongly disagree” and 5 meant “strongly agree.” The statements were as follows: (1) “After reading/viewing news reports about the student-led protests over the China trade pact, I have thought about the consequences of the issue”; (2) “After reading/viewing news reports about the student-led protests over the China trade pact, I have worried about the consequences of the issue”; (3) “I often try to relate what I see in the news about the student-led protests over the China trade pact to my personal experience”; and (4) “I often think about how what I see in the news about the student-led protests over the China trade pact relates to other things I know.” Principal components analysis showed that the four items grouped in a single factor, indicating that they measured a single underlying concept (Eigenvalue = 2.75, accounting for 68.75% of the variance). The four items were averaged to create a composite measure of “elaboration of the student protest news” (M = 3.78, SD = 0.63, α = .79).

Munroe (2002) defined political participation as the degree to which citizens are exercising their right to engage in such activities as to protest, to speak freely, to vote, or to get more energetically involved. Although no universal definition of political participation exits in the literature, the essence of political participation, as Lamprianou (2013) argued, is the “act of democratic citizenship” (p. 1). Accordingly, we defined political participation operationally as discussions about and participation in the student protests.

Online discussions of the protests

Using a 5-point scale (1 = “never” and 5 = “always”), respondents were asked to indicate how often they discussed the student-led protests over the China trade pact online as well as on social media sites such as Facebook, YouTube, and Twitter. Results of a principal components factor analysis yielded a single-factor solution (Eigenvalue = 2.80%, accounting for 70% of the variance). A composite scale of “online discussions of the protests” was built by averaging the four items (M = 3.28, SD = 1.03, α = .86).

Participation in the protests

Respondents were further asked to indicate whether they had (1) “participated in the student-led protests over the China trade pact”; (2) “asked classmates to support the student-led protests over the China trade pact”; and (3) “asked family members to support the student-led protests over the China trade pact.” The response categories were as follows: 1 = yes and 0 = no. The three items were summed to yield an index of “participation in the protests.” The index ranged from 0 to 3 for each respondent (M = 0.66, SD = 1.02, Kuder–Richardson Formula 20 [KR-20] = .77).²

Control variables

Respondents were asked about their gender, age, and political party affiliation. Respondents also reported the amount of time per day they spent reading newspapers, watching television news, and surfing online news. These variables were used as controls in regression analyses because previous studies indicated that they were related to perceived issue importance, news attention, news elaboration, or the third-person effect (Andsager & White, 2007; Gunther, 1995; Lo & Chang, 2006).

Results

H1 predicted that respondents would perceive news about the student protests to have a greater effect on others than on themselves. Table 1 presents the results of paired t tests, which provided support for the baseline third-person effect for both individual measures and the combined effects index, t(1, 132) = 13.33 at the p < .001. As expected, respondents perceived that others were more influenced by the news about the student protests than themselves. H1 was supported.

Table 1.

Mean Estimates of Perceived Effects of News Reports About the Sunflower Movement on Oneself and Others (n = 1,133).

Samples	Self	Others	t value
Attitudes toward the protests	3.07 (1.10)	3.59 (0.92)	15.24***
Attitudes toward the government	3.32 (1.16)	3.61 (0.98)	8.28***
Attitudes toward the police	3.20 (1.14)	3.65 (0.96)	12.55***
Attitudes toward the trade pact	3.21 (1.22)	3.58 (0.99)	10.06***
All	3.20 (1.00)	3.61 (0.87)	13.33***

Note. Standard deviations are in parentheses.

***

p < .001.

H2a and H2b predicted a positive relationship between perceived issue importance, news attention, and news elaboration. To test them, partial correlations were used. The results showed that perceived issue importance was significantly and positively related to news attention (r = .51, p < .001) and news elaboration (r = .46, p < .001) after controlling for gender, age, and political affiliation. H2a and H2b were supported.

H2c predicted that perceived issue importance about the student protests would be positively related to perceived effects of such news on oneself and others. To test it, two hierarchical regression analyses were performed. As Table 2 shows, perceived issue importance turned out to be the strongest predictor of the perceived effects of news about the student protests on oneself (β = .29, p < .001). In addition, it was also a significant predictor of perceived effects on others (β = .16, p < .001). These results indicate that the higher the perceived issue importance of news about the student-led protests over the China trade pact, the greater the perceived effects of such news on oneself and others. H2c was supported.

Table 2.

Hierarchical Regression Analysis Predicting Perceived Effects on Self and Perceived Effects on Others (n = 1,046).

Independent variables	Perceived effects on self	Perceived effects on others
Block 1: Demographics
Gender	.03	.04
Age	−.03	−.00
Party affiliation	−.15***	−.06*
Adjusted R²	5.7%	1.1%
Block 2: News exposure
Newspaper	−.02	−.04
TV news	.03	.05
Online news	.04	−.04
Incremental adjusted R²	4.3%	0.4%
Block 3: Information processing
Perceived issue importance	.29***	.16***
Attention	.20***	.10**
Elaboration	.16***	.13***
Incremental adjusted R²	22.3%	8.1%
Total adjusted R²	32.3%	9.6%

Note. Cell entries are standardized regression coefficients from the final regression equation with all blocks of variables in the model.

p < .05. **p < .01. ***p < .001.

H3 predicted that attention to the student protest news would be positively related to perceived effects of such news on oneself and others. To test it, two hierarchical regression analyses were performed. As Table 2 shows, news attention was significantly related to the perceived effects of news about the student protests on oneself (β = .20, p < .001) and perceived effects on others (β = .10, p < .01). H3 was supported.

H4 predicted that elaboration of the student protest news would be positively related to perceived effects of such news on oneself and others. Two hierarchical regression analyses were performed to test it. As shown in Table 2, news elaboration was a significant predictor of the perceived effects on oneself (β = .16, p < .001) and perceived effects on others (β = .13, p < .001). These results indicate that as news elaboration increases, the perceived news effects on oneself and others also increase. H4 was supported.

H5a predicted that perceived effects of the student protest news on oneself would be more strongly and positively related to online discussions of the protests than would be perceived effects of such news on others. To test it, a separate hierarchical regression analysis was performed. Table 3 summarizes the results (see column 1 in Table 3), which show that perceived effects on oneself were a significant and positive predictor of discussing the protests (β = .22, p < .001), whereas perceived effects on others were a weak and negative predictor (β = −.13, p < .001). Additional tests for the difference between two regression coefficients using Cohen and Cohen’s (1983) method show that the difference between the betas of perceived effects of student protests on the self and on others was significant (t = 7.16, p < .001).³ H5a was supported.

Table 3.

Hierarchical Regression Analysis Predicting Political Participation (n = 1,034).

Independent variables	Online discussions	Participation in the protests
Block 1: Demographics
Gender	−.04	.01
Age	.04	.02
Party affiliation	−.09**	−.16***
Adjusted R²	5.1%	7.1%
Block 2: News exposure
Newspaper	.02	−.01
TV news	−.05	−.03
Online news	.16***	.09**
Incremental adjusted R²	10.7%	4.7%
Block 3: Information processing
Perceived issue importance	.12***	.13***
Attention	.24***	.20***
Elaboration	.13***	.05
Incremental adjusted R²	17.4%	12.5%
Block 4: Perceived effects
Self	.22***	.24***
Others	−.13***	−.12**
Incremental adjusted R²	3.4%	3.8%
Total adjusted R²	36.6%	28.1%

Note. Cell entries are standardized regression coefficients from the final regression equation with all blocks of variables in the model.

p < .05. **p < .01. ***p < .001.

H5b predicted that perceived effects of the student protest news on oneself would be more strongly and positively related to participation in the protest than would be perceived effects of such news on others. Results of additional hierarchical regression analysis also show (see column 2 in Table 3) that perceived effects on oneself were positively and significantly associated with participation in the protests (β = .24, p < .001), whereas perceived effects on others were a negative predictor (β = −.12, p < .01). The difference between the betas of perceived effects of news about the student protests on oneself and perceived effects of such news on others was significant (t = 7.36, p < .001). H5b was supported.

It is important to note that perceived effects on others were a negative predictor of the two outcome variables. It seems that respondents who believed the protest news would affect others were not likely to take action in support of the protests. This pattern is consistent with the earlier studies of voting behavior that found the greater perceived effects on others, the less likely one is to vote (Banning, 2006). As we anticipated, perceived effects on the self, rather than on others, were a motivation to take action.

Finally, to address RQ1, which explored the mediating role of news information-processing variables and perceived effects of the protest news between perceived issue importance and political participation, we implemented structural equation modeling using Amos 18.0. In the model, perceived effects on others were excluded because it was weakly and negatively related to both online discussions of the protest and offline participation in the protests.

The results of the analyses showed that although the chi-square for the model was significant, χ² = 742.10, df = 166, p < .001 (χ²/df ratio = 4.47), the comparative fit index (CFI = .95), the normal fit index (NFI = .94), the Tucker–Lewis index (TLI = .93), and the root mean square error of approximation (RMSEA = .055) indicate that the model fit was acceptable. The model explained 29.3% of the variance in news attention, 49.0% in news elaboration, 37.2% in perceived effects on oneself, 33.1% in participation in the protests, and 27.5% in online discussions.

As shown in Figure 2, the structural equation model indicates that perceived issue importance had a significant effect on news attention (β =.54, p < .001), which was significantly associated with news elaboration (β = .49, p < .001), perceived effects on oneself (β = .29, p < .001), and online discussions of (β = .28, p < .001) and participation in the protests (β = .26, p < .001). Perceived issue importance also had a significant effect on news elaboration (β = .30, p < .001), perceived effects on oneself (β = .35, p < .001), and participation in the protests (β = .14, p < .01). The structural equation model further affirms that perceived effect on the self was a significant predictor on the two endogenous variables: online discussions of the protests (β = .17, p < .001) and participation in the protests (β = .20, p < .001). Although news elaboration significantly enhanced online discussions (β = .18, p < .001), it did not significantly relate to perceived effects on oneself (β = .07, p > .05) and participation (β = .04, p > .05).⁴

Figure 2.

The structural equation model.

To assess the mediating effects of news processing variables and perceived effect on oneself in enhancing the effects of perceived issue importance on political participation, we adopted a procedure developed by Sobel (1982) that provides a direct test of an indirect effect. The three mediators in our theoretical model are news attention, news elaboration, and perceived effects on self. Thus, there are three potential mediating effects between perceived issue importance and online discussions of the protests in the model. The results of the Sobel test show that the z score for mediation path through news attention is 4.81 (p < .001), the path through news elaboration has a z score of 0.76 (p > .05), and the path through perceived effects on self has a z score of 4.42 (p < .001). Therefore, we generated evidence in support of the proposed model that news attention and perceived effects on self are statistically significant mediators in the relationship between perceived issue importance and online discussions of the protests.

Similarly, news attention, news elaboration, and perceived effects on self are also three potential mediators in the relationship between perceived issue importance and participation in the protests. The results of the Sobel test indicate that the mediation path through news attention has a z score of 5.27 (p < .001), the z score for the path through news elaboration is 3.96 (p < .01), and the z score for the path through perceived effects on self is 3.46 (p < .01). Accordingly, news attention, news elaboration, and perceived effects on self are significant mediators in the relationship between perceived issue importance and participation in the protests.

An additional bootstrapping procedure was conducted using the SPSS version of PROCESS macro developed by Hayes (2013). The bootstrap method demonstrates that the mediation effects we reported with the Sobel test were accurate and appropriate given the nature of our data. Specifically, the bootstrapping procedure was used with 5,000 bootstrap samples and 95% bias-corrected bootstrap confidence intervals. The results are summarized in Tables 4 and 5.

Table 4.

Indirect Effects of Perceived Issue Importance on Online Discussion Through the Hypothesized Mediators.

Variable	B	SE (boot)	k² [CI]
Attention	.208	.021	.21 [.17, .25]
Elaboration	.187	.023	.19 [.14, .24]
Perceived effects on self	.203	.026	.20 [.16, .26]

Note. Five thousand bootstrap samples were used to generate 95% bias-corrected CIs. Values listed for k² represent the ratio of the observed indirect effect relative to the maximum possible indirect effect. In this case, k² of .10 simply means the indirect effect through the mediator is 10% of the total possible indirect effect that could hypothetically occur through said mediator. The CIs associated with k² show the magnitude of the effect. Adopting Cohen’s guidelines, the effect size can be interpreted as small if the CIs are entirely above .01, medium if they are entirely above .09, and large if they are entirely above .25. CI = confidence interval.

Table 5.

Indirect Effects of Perceived Issue Importance on Participation in the Protests Through the Hypothesized Mediators.

Variable	B	SE (boot)	k² [CI]
Attention	.161	.020	.17 [.13, .21]
Elaboration	.111	.022	.12 [.07, .16]
Perceived effects on self	.193	.024	.19 [.15, .24]

Note. Five thousand bootstrap samples were used to generate 95% bias-corrected CIs. CI = confidence interval.

Discussion

A controversial trade pact with China caused widespread student-led protests in Taiwan. The present study sought to examine how news coverage of the protests and perceived importance of the protest movement affected people’s evaluation of media’s impact on themselves and others. The results provided strong support for the third-person hypothesis. As expected, respondents perceived the protest news to have a greater influence on others than on themselves.

Given students’ fear that the pact would mean high unemployment, the results of the study further revealed that perceived importance of the protest movement was a significant predictor of perceived effects of the protest news on oneself and others. The more the student-led protests were perceived to be an important issue, the greater the perceived effects of such news on oneself and others. In addition, news attention and news elaboration were found to be significantly and positively related to perceived effects on oneself and on others. That is, the more the respondents paid attention to and elaborated on the protest news, the greater impact they perceived such news on themselves and others. It seems that respondents internalized the protest news. A plausible agenda-setting effect comes to the forefront in making sense of these findings because the salience of the protests in the news appears to have caused people to think about the protests. The more they thought about it, the greater effect they perceived on themselves.

Another goal of the study was to explore how perceived effects of news coverage of the student-led protests affected people’s political behavior. The results demonstrated that perceived effects of protest news on oneself were positively and significantly related to participation, whereas perceived effects on others were a significant but negative predictor.

These findings suggest that when news about the protests was perceived as having a personal impact, those who were politically involved were motivated to participate in the protests. When the protest news was perceived as having an impact on others, people were less likely to participate. Past research on the influence of presumed influence (Gunther & Storey, 2003) suggested that people’s perception of media effects on other people resulted in attitudinal and behavioral changes. We demonstrate that it was the perceived effect of protest news on oneself that prompted people to participate in the protests. These are theoretically important findings because they help advance the influence of presumed influence and third-person effect research by showing that perceived effects on the self and perceived effects on others play different roles in triggering political behavior. Consistent with Golan and Day’s (2008) proposition, perceived effects of political messages on the self were indeed a better predictor of political action than perceived effects on others.

Perhaps the most important findings of this study are that information-processing variables together with perceived effects on the self seem to mediate the relationship between perceived issue importance and political participation. Past research suggests that news attention and news elaboration mediate the relationship between motivation and political knowledge (Eveland, 2002). The results of the structural equation model indicate that both news attention and news elaboration play a mediating role between perceived issue importance and participation. Similarly, perceived effects on oneself were also a significant mediator between perceived issue importance and political participation. These findings indicated that those who perceived the student-led protests over the China trade pact to be personally important would pay greater attention to and elaborate on the protest news, and thus perceived such news as having a greater impact on themselves. Consequently, they were more likely to participate in the protests. On the contrary, those who did not participate in the protests did not bother to think about it. These findings shed some light on how news coverage of the protest movement and perceived issue importance of the movement affected political participation. In practical terms, it was probably the fear of devastating effects of the trade pact on employment that sent college students into the street. The salient news reports of the protests inspired more students to participate in the protests to protect their self-interest.

To balance the conclusions, a few limitations are acknowledged. Use of a sample drawn from college students limits the generalizability of the findings to the general public. Hence, the sample used in this study is a major limitation. The findings should be interpreted within the context of this study of participation in a student-led movement. Student movements can have a profound impact on the general population and on a nation’s politics; the anti-Vietnam war movement in the United States and Taiwan’s previous moments such as Wild Lily and Wild Strawberries are examples. In this sense, the Sunflower Movement is similar in influencing public opinion about and government policy toward China. The China pact was suspended as a result of the protests.

Another limitation of this study is that the measure of political participation did not include all voluntary activities that might be taken by students to influence the outcomes of the protest movement such as sign petitions, donate money, contact officials or politicians, and attend political meetings. Future research should include more voluntary activities. In addition, this study measured online discussions of the protests without further measuring offline interpersonal discussions of the protests. This is yet one more limitation because past research suggests that interpersonal discussion is related to information processing and political knowledge (Eveland, 2004; Robinson & Levy, 1986). It would thus be instructive to explore how perceived effects influence offline discussions.

Also, the relationships reported are not necessarily causal. It is plausible that perceived issue importance plays a mediating role between information-processing variables and perceived effects. That is, paying attention to and elaborating on protest news may lead individuals to perceive the protest movement to be personally important, which in turn may affect perceived effects of such news on themselves and others. Future research can explore how perceived issue importance mediates the relationship between information-processing variables and the third-person effect. Finally, Taiwanese students may differ from mainland students in perceiving the influence of news about the movement on themselves and others because mainland students are more likely to support the trade pact with China and less likely to participate in the protests. However, this study did not measure the students’ home residence. Examining the impact of news on the third-person effect and political participation between different groups or across cultures would be an important avenue for future research.

Footnotes

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) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Ven-Hwei Lo (PhD, University of Missouri–Columbia) is a professor in the School of Journalism and Communication at the Chinese University of Hong Kong.

Ran Wei (PhD, Indiana University, 1995) is Gonzales Brothers Professor of Journalism in the School of Journalism and Mass Communications at the University of South Carolina.

Hung-Yi Lu (PhD, University of Kentucky) is a professor in the Department of Communication and Graduate Institute of Telecommunication, National Chung Cheng University, Taiwan.

References

Andsager

J. L.

White

H. A.

(2007). Self versus others: Media, messages, and the third-person effect. Mahwah, NJ: Lawrence Erlbaum Associates.

Banning

S. A.

(2006). Third-person effects on political participation. Journalism & Mass Communication Quarterly, 83, 785-800. doi:10.1177/107769900608300404

Beaudoin

C. E.

Thorson

(2004). Testing the cognitive mediation model: The roles of news reliance and three gratifications sought. Communication Research, 31, 446-471. doi:10.1177/0093650204266098

Bryant

Miron

(2004). Theory and research in mass communication. Journal of Communication, 54, 662-704. doi:10.1111/j.1460-2466.2004.tb02650.x

Chang

(2014, April 23). The Sunflowers can take on Beijing. Taipei Times. Retrieved from http://www.taipeitimes.com/News/editorials/archives/2014/04/23/2003588663

Cohen

(1983). Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum Associates.

Cohen

Tsfati

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

Conge

(1988). The concept of political participation: Toward a definition. Comparative Politics, 20, 241-249.

Davison

W. P.

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

10.

de Zúñiga

Molyneux

Zheng

(2014). Social media, political expression, and political participation: Panel analysis of lagged and concurrent relationships. Journal of Communication, 64, 612-634. doi:10.1111/jcom.12103

11.

Eagly

A. H.

Chaiken

(1993). The psychology of attitudes. Fort Worth, TX: Harcourt.

12.

Eveland

W. P.

Jr. (2001). The cognitive mediation model of learning from the news evidence from non-election, off-year election, and presidential election contexts. Communication Research, 28, 571-601. doi:10.1177/009365001028005001

13.

Eveland

W. P.

Jr. (2002). News information processing as mediator between motivations and public affairs knowledge. Journalism & Mass Communication Quarterly, 79, 26-40. doi:10.1177/107769900207900103

14.

Eveland

W. P.

Jr. (2004). The effect of political discussion in producing informed citizens: The roles of information, motivation, and elaboration. Political Communication, 21, 177-193. doi:10.1080/10584600490443877

15.

Eveland

W. P.

Jr. Shah

D. V.

Kwak

(2003). Assessing causality in the cognitive mediation model: A panel study of motivations, information processing, and learning during campaign 2000. Communication Research, 30, 359-386. doi:10.1177/0093650203253369

16.

Fung

F.W.Y.

(2014, March 30). Ma yields on pacts ahead of rally. South China Morning Post. Retrieved from http://www.scmp.com/news/china/

17.

Golan

G. J.

Day

A. G.

(2008). The third-person effect and its behavioral consequences: A new trend in twenty-five-year history of third-person effect research. Mass Communication and Society, 11, 539-556. doi:10.1080/15205430802368621

18.

Griffin

R. J.

Dunwoody

Neuwirth

(1999). Proposed model of the relationship of risk information seeking and processing to the development of preventive behaviors. Environmental Research, 80, S230-S245. doi:10.1006/enrs.1998.3940

19.

Gunther

A. C.

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

20.

Gunther

A. C.

Storey

J. D.

(2003). The influence of presumed influence. Journal of Communication, 53, 199-215. doi:10.1111/j.1460-2466.2003.tb02586.x

21.

Hayes

A. F.

(2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: The Guilford Press.

22.

Lamprianou

(2013). Contemporary political participation research: A critical assessment. In Demetrious

K. N.

(Ed.), Democracy in transition (pp. 21-42). Berlin, Germany: Springer.

23.

Lin

S.S.

(2015, December 10). Sunflowers and umbrellas: Government responses to student-led protests in Taiwan and Hong Kong. The Asan Forum. Retrieved from http://www.theasanforum.org/sunflowers-and-umbrellas-government-responses-to-student-led-protests-in-taiwan-and-hong-kong/

24.

Liu

V. H.

(2014). Media exposure, perceived personal impact, and third-person effect. Media Psychology, 17, 378-396. doi:10.1080/15213269.2013.826587

25.

V. H.

Chang

C. C.

(2006). Knowledge about the Gulf wars: A theoretical model of learning from the news. The Harvard International Journal of Press/Politics, 11, 135-155. doi:10.1177/1081180X06289582

26.

V. H.

Wei

(2002). Third-person effect, gender, and pornography on the Internet. Journal of Broadcasting & Electronic Media, 46, 13–33. doi:10.1207/s15506878jobem4601_2

27.

V. H.

Wei

H. Y.

Hou

H. Y.

(2015). Perceived issue importance, information-processing and third-person effects of news about imported U.S. beef controversy. International Journal of Public Opinion Research, 27, 341-360. doi:10.1093/ijpor/edu039

28.

Munroe

(2002). An introduction to politics: Lectures for first-year students. Kingston, Jamaica: Canoe Press.

29.

Neuwirth

Frederick

Mayo

(2002). Person-effects and heuristic-systematic processing. Communication Research, 29, 320-359. doi:10.1177/0093650202029003005

30.

Perloff

R. M.

(1993). Third-person effect research 1983–1992: A review and synthesis. International Journal of Public Opinion Research, 5, 167-184. doi:10.1093/ijpor/5.2.167

31.

Perloff

R. M.

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

32.

Perloff

R. M.

(2009). Mass media, social perception, and the third-person effect. In Bryant

Oliver

M. B.

(Eds.), Media effects: Advances in theory and research (3rd ed., pp. 252-268). New York, NY: Routledge.

33.

Petty

R. E.

Brinol

Priester

J. R.

(2009). Mass media attitude change. In Bryant

Oliver

M. B.

(Eds.), Media effects: Advances in theory and research (pp. 125-164). New York, NY: Routledge.

34.

Petty

R. E.

Cacioppo

J. T.

(1979). Issue involvement can increase or decrease persuasion by enhancing message-relevant cognitive responses. Journal of Personality and Social Psychology, 37, 1915-1926.

35.

Petty

R. E.

Cacioppo

J. T.

(1986). Communication and persuasion: Central and peripheral route to attitude change. New York, NY: Springer-Verlag.

36.

Price

Stroud

J. N.

(2005). Public attitudes toward polls: Evidence from the 2000 US presidential election. International Journal of Public Opinion Research, 18, 393-421. doi:10.1093/ijpor/edh119

37.

Robinson

J. P.

Levy

M. R.

(1986). Interpersonal communication and news comprehension. Public Opinion Quarterly, 50, 160-175.

38.

Rojas

(2010). “Corrective” actions in the public sphere: How perceptions of media and media effects shape political behaviors. International Journal of Public Opinion Research, 22, 343-363. doi:10.1093/ijpor/edq018

39.

Rojas

Puig-i-Abril

(2009). Mobilizers mobilized: Information, expression, mobilization and participation in the digital age. Journal of Computer-Mediated Communication, 14, 902-927. doi:10.1111/j.1083-6101.2009.01475.x

40.

Shaw

Hamm

Terry

(2006). Vertical versus horizontal media: Use agenda-setting and audience agenda-melding to create public information strategies in the emerging papyrus society. Military Review, 86, 13-25.

41.

Sobel

M. E.

(1982). Asymptotic intervals for indirect effects in structural equation models. In Leinhart

(Ed.), Sociological methodology (pp. 290-312). San Francisco, CA: Jossey-Bass.

42.

Wei

V. H.

(2015). The third-person effect. In Mazzolen

(Ed.), The international encyclopedia of political communication (pp.1-6). London: Blackwell.

43.

Wei

V. H.

H. Y.

(2007). Reconsider the relationship between the third-person effect and optimistic bias. Communication Research, 34, 665-684. doi:10.1177/0093650207307903

44.

Wei

V. H.

H. Y.

(2010). The third-person effect of tainted food product recall news: Examining the role of credibility, attention, and elaboration for college students in Taiwan. Journalism & Mass Communication Quarterly, 87, 598-614. doi:10.1177/107769901008700310

45.

Witte

(1994). Fear control and danger control: A test of the extended parallel process model (EPPM). Communication Monographs, 61, 113-134. doi:10.1080/03637759409376328

46.

Witte

Morrison

(2000). Examining the influence of trait anxiety/repression-sensitization on individuals’ reactions to fear appeals. Western Journal of Communication, 64, 1-27. doi:10.1080/10570310009374661