Gender Stereotypes and Citizen Journalism

Every citizen is a reporter, Oh Yeon-ho said as he launched Ohmynews in 2004, largelyas a fierce reaction to partisan politics in South Korean (Oh, 2004). “The citizens of the Republicof Korea had long been preparing for a grand revolution in the culture of news production andconsumption,” he said. “All I had to do was raise the flag.” The success of his site provided ahigh profile example of a movement many others had predicted. Glaser defined citizen journalism as the idea that “people without professional journalism training can use the tools of modern technology and the global distribution of the Internet to create, augment or fact-check media on their own or in collaboration with others.” (2004). Gillmor (2004) applauded the abilityof the “people formerly known as the audience” to bypass legacy news organizations, such as TVnetworks, newspapers, and magazines, and report the news their own way.

When the citizen journalism movement reached the United States, however, it took on amuch different flavor than the fierce political discussion of Ohmynews. The Northwest Voice, acitizen journalism website in Bakersfield, Calif., began in early 2004 as well, and became one of the first citizen journalism experiments in the United States. Its content, however, focused moreon human-interest stories (Glaser, 2004). Some of the site’s most viewed stories covered lighter topics, such as a mother’s reaction to sending her child to his first day of school (Meyer, 2006). Other citizen journalism sites followed a similar script.

In a case study of MyMissourian.com, aColumbia, Mo. citizen journalism website based on the Northwest Voice, Bentley et. al (2007)found that the most popular stories covered issues that normally did not receive much attention in the mainstream media. Women, for example, filled the site with stories of their lives andcauses, even though the site was originally created to discuss the 2004 presidential elections (pg.249).The idea that citizen journalism exists to cover only issues of the fairer sex, however, is amisconception. Mayhill Fowler, for example, has bucked the trend since 2008.

The Huffington Post’s

Off the Bus project that contained influential information about what politicalfigures said in her presence not knowing she was on the beat as a reporter (Boehlert, 2009). Herboss Arianna Huffington is herself an influential figure in citizen journalism. “At HuffPost, wesee citizen journalism as an integral part of what we do – and, via Eyes and Ears, our citizen journalism community, we’re harnessing the wisdom of the crowd to tackle stories too big forone reporter, attend events traditional journalists have been kept from (or have overlooked), andto find and highlight the small but evocative story happening right next door,” she wrote (2009).Another citizen journalist argued the movement needs more female voices to flourish.“If citizen journalism is to make the strides it should … there needs to be parity of the sexes and a reliable sense of mutual respect. Otherwise it might as well be tabloidwriting. Muting of female voices will only be a detriment to the concepts behind citizen journalism” (Askcherlock, 2010).

For citizen journalism to make the gender equality strides Askcherlock mentions, media professionals and researchers need to better understand how audience members of both gendersreact to user-generated content. They also should determine what role, if any, gender plays in credibility and some of its core components, such as social presence, coorientation, andexpertise.This is an important question to consider as women have closed the gap in overallInternet use (Pew, 2012). More than 85 % of men and women say they use the Web regularly. Citizen journalism opportunities are also expanding. Dube (2010) lists 23 citizen mediainitiatives created by legacy media organizations, including CNN’s iReport, CBS’ EyeMedia,BBC’s iCan, and independent sites such as MapYourMoments. CNN paid more than $750,000for the iReport domain name in 2008 to expand its ability to publish user-generated content(Learmonth, 2008).

This study uses an online experiment to examine if participants’ reactions to stories published on a major news organization website differ based on the story’s author and his or her declared gender. The study has a specific focus on credibility and the measures that lead up to it,such as social presence, coorientation, and expertise, and asks whether, if at all, men and womenreact differently to staff writers and citizen journalists, especially if the author’s gender matchestheirs. Social presence, for example, will answer whether participants notice authors that sharetheir gender more than others. Coorientation will help explain if participants think authors sharetheir beliefs and perceptions more if they also share their gender. Expertise, as a concept, willhelp explain whether participants think authors that share their gender are more of an expert onthe topic they are covering and therefore more trustworthy. As this study deals with audience perceptions based on gender cues, it could also help inform why some authors would choose tohide or even alter their gender online to reach a certain audience.

Literature Review

Pearson (1982) was one of the first to suggest male writers have more credibility thanfemale writers. She also wrote in her book (1985) that female writers believed they would havemore credibility if they were male, and writers of both sexes thought they had more credibilitywith members of their own sex than with those of the opposite sex.White and Andsager (1991) also suggested that women found newspaper opinioncolumns written by women more interesting than those written by men, while men felt the sameabout male-written columns. However, they suggested that there would be no difference incredibility between male and female writers regardless of the reader’s gender. In other words, thetype of story, they suggested, mattered.

News pieces seem more credible on the surface thanopinion columns.These credibility differences between men and women revolve around basic gender stereotypes (Deaux & Lewis, 1984). A gender label alone can lead people to infer a variety of gender-related characteristics (pg. 1002). Several studies (Broverman, Vogel, Broverman, Clarkson, & Rosenkrantz, 1972; Rosenkrantz, Vogel, Bee, Broverman, & Broverman, 1968)identified two sets of characteristics that are ascribed to women and men. Subjects associatedwarmth and expressiveness with women more than men, and competence and rationality, withmen more than women. Personality traits are the essence of gender stereotypes (Deaux andLewis, 1984). Gender stereotypes are able to bind people together, and also are important in thesocial perception process (Grant and Holmes, 1981, 1982).

Deaux and Lewis (1984) also found information about how one stereotype componentcan influence other components. However, gender information alone does not lead to theinference of a certain set of characteristics. “In most cases, however, the influence of gender can be outweighed by other information, such as role behaviors, traits, and the like” (pg. 1002). Inother words, if readers only have the author’s name and lack other cues about the author’scredibility, they may fall on gender stereotypes.

These same gender stereotypes seem to drive how men and women use a medium. Menlook for information on a wider variety of topics than women do, “from researching products to buy to getting information on their hobbies to looking for political news. Sometimes, men andwomen look for different kinds of information” (Fallows, 2005). After the events of September 11, men visited more websites to tell them about things that were happening; more women saidthe internet helped them find people they needed to reach. (p.iv). A larger percentage of womenthan men said in the study that they suffered from information overload online. While men andwomen both appreciate what the Internet does for their lives, men said they valued how it helpstheir activities, such as jobs and pastimes, more than women while women valued relationshipswith family, friends, colleagues, and communities (pg. v).

Both genders are increasingly exposed to cues beyond gender stereotypes to help themdetermine the credibility of the information they find online. These other cues can mitigate theeffect of gender stereotypes. Matheson (1991) found that within a computer-mediatedcommunication environment, participants who thought they were interacting with a womanfound her to be more cooperative and less exploitive. This occurred only when participants wereexplicitly told the person on the other side of the conversation was a woman. “It is conceivablethat such information focused subjects on the characteristics of the other that were similar totheir own, and hence invoked gender relevant dimensions of themselves, which were, in turn,‘projected’ onto the negotiator” (pg. 143). It took a strong suggestion to outweigh other cues andcause participants to go back to their default understanding of gender.

Flanagin and Metzger (2003) suggested sex is a significant factor when looking at perceived website credibility measures. “It is the interaction between author’s and viewer’s sexthat seems at the base of this relation” (pg. 698). For instance, their research suggested same-sexcredibility evaluations were lower than opposite sex credibility evaluations. Credibility would behigher when the sex of the message source matched the sex of the receiver. Four years later,Flanagin and Metzger (2007) underscored the importance of website features in determining perceived credibility. They suggested that a site’s frame, such as whether it was a news or shopping site, was the main determination of credibility. Their earlier study looked only at whatthey called “personal websites,” where a single author was clearly identified and she, in this case,was sharing about her life. For instance, participated said reviews on shopping websites from people like them were more credible than expert reviews. On news websites, however, expertopinions were rated the most credible.

Armstrong and McAdams (2009) supported these findings when they suggested thatinformation-seeking was one of the main determinants of credibility. Their study focused onWeblogs or blogs, one of the most common user-generated content forums online. A blog allowsan author to publish her own news stories or thoughts, feelings, or opinions in a chronologicalformat. A blog is primarily a content delivery system. The authors manipulated the gender descriptors of the blog’s author, and found that gender cues can influence the perceivedcredibility of blogs because individuals may perceive some topics as “belonging” to female or male bloggers or as requiring a particular expertise. They found blog posts written by men weredeemed more credible than those written by women, but the writing style and topic of the blogwere also likely to influence the perceived credibility of the post. As blogs become more prevalent and accepted as sources of credible information, they suggested that gender cues wouldbecome less important. Tone, in fact, could have a greater influence than gender. A more cynicaltone increased perceived credibility among young people.

Credibility

To understand the effect of gender on credibility, however, it is important to define theconcept more precisely for an Internet age. Its definition is not as straightforward as it was 10years ago. It is still true that the more credible people find a news source, the more likely theywill be to use it. However, people use media that they say are not credible. For example, Reeves& Nass (1996) found audiences consistently awarded higher credibility marks to TV, eventhough TV stories lack the depth of and borrow extensively from newspaper reports. The reasonthey suggested was the human dimension.

At its most basic level, credibility is not simply an objective measure of a medium’sfeatures or messages (Flanagin & Metzger, 2007). It revolves around subjective evaluations of how stories, sources and organizations are presented. Reeves & Nass (1996), Wackman (1973), Kim (2010), Sundar (1999) and Rafaeli (1988) have shown that credibility needs to includemeasures of how much a person likes a medium, how much they have come to rely on it, andhow connected they feel to it and its agents.

The human dimension of the “credibility crisis” Gaziano first wrote about in 1986 has become even more complicated in 2012. The Internet has made more information available thanever, while muddling some of the core concepts of traditional credibility. Deciding what iscredible requires examining relationships between people and their sources for news even more.

Early media scholars approached credibility from two measures: do you trust the mediaand do you believe what you read? Gaziano & McGrath (1986) expanded the definition toinclude 12 measures. Their scale coupled questions about trust and believability with concepts such as objectivity, complexity, completeness, truth and reputation. Despite the comprehensivenature of the Gaziano & McGrath scale, researchers found other elements. Meyer (1988), aformer newspaper editor himself, boiled the Gaziano & McGrath scale down to one element – believability – while adding the idea of community affiliation. Beaudoin & Thorson (2004)reinforced Meyer’s addition by suggesting that credibility grows the more the newspaper connects to the community. Perloff (2003) added another human element in examining persuasive research – perceived expertise. Expertise links credibility to what people judge as theextent of a source’s knowledge and experience on a topic.

Another way to look at credibility is to break it into separate components that examinedmessage, source, and organizational credibility separately. Source credibility, for example, couldinclude Perloff’s expertise dimension, while organizational credibility spoke more to Meyer’saddition of affiliation. Separating the concepts has proven useful. Sundar (1999) demonstratedthe need to examine source credibility on its own when he asked readers to rate stories based onthe type and number of quotes they had. But he also found an interesting connection between a person’s relationship with a source and credibility. People judged sources based more on who thesource was rather than what he or she said. His study speaks to the personal nature of credibilitydefinitions, whether they deal with sources, messages or organizations. The human elements of messages take three forms – social presence, coorientation, and expertise – which work together to create a connection that leads to credibility and override gender stereotypes.

Connection

Social Presence

The ways in which non-human agents such as TV news broadcasts or even newspapersmake receivers feel as if they are human constitutes what researchers call social presence. The three dimensions social presence are (1) source attention, or how much focus the source receivesin the presentation relative to other cues, (2) co-presence, or how much an audience member canfeel the other person’s existence, and (3) mutual awareness or the feeling of being “known” byanother (Biocca et al., 2001; Gunawardena & Zittle, 1997; Tamborini & Skalski, 2005). Social presence is the personal characteristics that make a receiver connect with the source. It can becreated from pictures, word choice, and tone, among other elements.

Coorientation

Recognizing a human presence is not enough, however. Credibility also depends on howclosely a person allies with the source. Wackman (1973) wrote the goal of communication for information exchange is to increase coorientation between two people. He defined coorientationas the level of similar attitudes and the perceived congruency of those attitudes. Findingsomething to relate to in the media is not hard. In fact, Reeves & Nass (1996) said it is natural.

Inwhat they called the “media equation,” they suggested coorientation between source and receiver existed even as messages were disseminated through the media. People have evolved to respondto other humans. When they see something that resembles a human, they respond to the mediathe same way they would to another person, and they tend to like media that act in a humanfashion. Gender stereotypes play a role in how audience relate to the media.

Expertise

Another element of source credibility is perceived expertise. Hovland et al. (1953)suggested two dimensions of source credibility: trustworthiness and expertise. They argued thata receiver’s tendency to accept a speaker’s message depends on her estimation of how informedand intelligent the speaker is and how likely the speaker is to make valid points. Perloff (2003)defined expertise as a core characteristic of credible communicators. Whether a communicator should emphasize her expertise or her similarity to the audience can be a dilemma. Stories onfactual matters, such as news, might rely more on an expert’s knowledge than similarity (Perloff,2003). In order for news to connect with readers, it must pay attention to each of the threeelements – social presence, coorientation, and expertise – because stronger connections have theability to dramatically alter perceptions. As connection grows, so does credibility. Sundar (1999)identified relatedness as one of four elements that defined both online and newspaper credibility.

Greater social presence can also lead to reliance, which Wanta (1994) suggested can increasecredibility.Defining credibility for the Internet age requires more than just a study of site features.Flanagin & Metzger (2007) found that source matters. The nature of the organization and itsmessages were determining factors in the credibility respondents assigned. Participantsconsistently gave news sites the highest credibility scores, even when they had never seen that particular news site before. They could learn from the social cues the site provided how credibleit was, but interestingly, the more social presence the site had the less credible it was. Personalweblogs were rated the least credible even when they had the exact same, almost verbatimstories, as the news, and e-commerce sites. Meyer, Marchionni & Thorson (2006) also foundsocial presence was not positively related to the credibility of news websites. The main predictor in their study was expertise. Subsequent analysis revealed coorientation not social presence positively predicted expertise.

Credibility online

The impact of new credibility concepts, such as coorientation, social presence, expertiseand interactivity online does not invalidate the more traditional credibility definitions originally applied to newspapers. Credibility online must join the Web’s ability to make connectionsthrough technology with traditional concepts of trust, believability and expertise. Definingcredibility also necessitates an understanding of the values and purpose of Web communication.Sites that connected most with audiences (Flanagin & Metzger, 2007) are sites that presenttrustworthy information in formats that spoke to the audiences and allowed them to connect.Johnson & Kaye (2004) also suggest relationships help determine why users sometimesrate opinionated blogs more credible than news Web sites.

They linked credibility both on andoffline, with how familiar a person was with the medium. Those with more familiarity with theInternet and less with traditional media rated political blogs more credible. Traditional mediacould attract an audience with less familiarity with traditional news media online if they adoptedmore of the social presence and coorientation attributes of the Web.To measure and increase credibility in the Internet age, researchers need to examine notonly if trust, belief, and expertise exist. They must look at how individuals come to create thesefeelings and how the options available online can help. Most importantly, connection may enableus to overcome gender stereotypes in information processing.

Hypotheses and Methods

Based on the literature, this study examined the following hypotheses within the citizen journalist context:

H1.

Gender match, or whether the receiver’s gender matches that of the sources,will have a direct positive effect on credibility perceptions for stories writtenby a news organization staff writer.

H2.

Gender match will have a direct positive effect on credibility perceptions forstories written by audience members.

H3.

The components of credibility, such as coorientation, social presence, andinterest, will mitigate the effects of gender on credibility.

The study is based on a 2 (story author) x 2 (author’s gender) within-subjects experiment,which asked participants to read four news stories, all ostensibly coming from the same major news website. The stories were basic news stories on six separate topics selected to be non-controversial. The researchers specifically avoided stories dealing with politics, religion or women’s issues. The topics were technology, health, U.S. news, entertainment, world news, andcrime. The gender of the story author, along with whether the author was a staff writer or audience member was manipulated. In the end, each participant read a story written by a maleand female staff writer and a male and female audience member. The order of the authorshipconditions, the topics, and the writer’s gender were randomized through a simple website-based program called a PHP script.

The random assignment and within-subjects design helped theexperiment to focus on the variance authorship created, not other external factors.Respondents answered the same questions after each story. Respondents rated the social presence, expertise, coorientation, credibility, and interest they had in each story. The questionsfollowed Meyer, Marchionni & Thorson (2010), which used a similar model to determine how participants ranked the credibility of stories written in traditional news, opinionated, and citizen journalism formats. Story titles for this study included “After tornado, town rebuilds by goinggreen,” “Air Force One Backup Rattles New York Nerves,” “In digital age, can movie piracy bestopped?” and “Students, musicians fight and fear Taliban.”Social presence measured how much readers noticed the person behind the story, with questions such as “I felt like I got to know the author,” “At times, I felt like the author was in theroom with me,” and “I thought of the author while reading the article.” Expertise measured thelevel of proficiency participants thought the author had with the topic through questions such as (either a Master’s or a Ph.D.).

Another 53 % reported incomes less than $25,000 annually, while12 % had incomes of more than $100,000 a year.The story was the unit of analysis. Each participant then had six units of analysis.Statistical tests focused on the differences between the participant’s gender and whether itmatched the author’s. An independent samples T-test focused on all the stories, asking only if gender matched. The only statistical significance was found when an audience member was theauthor. Expertise (t(343)=-2.22, p<.01) and story credibility (t(343)=-2.66, p<.01) were bothstatistically significant in the negative direction. This meant that when the author’s gender didnot match the participant’s, the story was rated less expert and less credible. (See table 1). Next, the researchers examined if any difference existed for men and women across theauthorship conditions.

For men, the only statistically significant variable was story credibility(t(374)=2.62, p<.01) if the story was written by a staff member. Male participants (see table 3)found the staff written story more credible if the author’s gender did not match. In other words,male participants rated stories written by staff writers who were women more credible than thosewritten by men.For women (see table 4), statistical significance was found only for audience writtenstories. If the author’s gender did not match the participant’s, women found less social presence(t(343)=-.248, p<.01), story credibility (t(343)=-2.42, p<.01), and interest (t(343)=-.346, p<.01).In other words, women assigned more social presence, story credibility and interest to storieswritten by other women. These t-tests provided no support for:

H1:

Story credibility will dependon matching gender for staff written stories.

H2:

Credibility for audience written stories will also rely on gender match. For staff written stories, only menseemed to care about gender match and were likely to rate stories written by women more credible. For audience written stories, women rated stories written by other women morecredible, and also gave those stories higher social presence and coorientation scores.To determine the effects of the variables that the literature suggests can predict credibility,univariate ANOVAs examined if there was any interaction between the story’s author andgender match. For men, only interest was statistically significant in the ANOVA (see table 5), and gender match was the only influence.

For women, social presence, story credibility, and interest had statistical significance,while expertise, even though it was significant in the t test, had no statistical significance whenthe gender match and authorship were examined together (see table 6). Social presence was notsignificant with gender match alone, but was significant for the author condition and for theinteraction. Story credibility was significant for gender match and authorship, but not in theinteraction. Interest was significant for gender match and the interaction, but not for authorship.

H3:

The predictors of credibility will mitigate the impact of gender cues. The only element that affected credibility for the men when the author’s gender matched was interest, and this variable was not statistically significant for the authorshipcondition. In other words, interest determined credibility for men more than gender.

For women, story credibility was affected by both gender match and authorship, but notin the interaction. Authorship was statistically significant at the p<.01 level, while gender matchwas significant at the p<.05 level. Women related better to stories, staff or audience written, that present their point of view, but find staff written stories more credible. However, interactioneffects were found for social presence and interest, which suggest these variables mediate gender effects. Women noticed other women more in audience written stories, while they gave higher interest scores to stories written by women, regardless of whether the author was a staff writersor an audience member.

Discussion

These findings suggest that a more nuanced relationship between gender and credibility iscontinuing to emerge online. This study examined the growing citizen journalism movement todetermine whether audience written stories are more susceptible to gender stereotypes. This wasan effort to explore ways to encourage women to participate in citizen journalism, regardless of story content. As Mayhill Fowler has demonstrated, women citizen journalists are just as capableof covering politics as they are of chronicling their child’s first day of school.What this study suggests is that traditional journalistic definitions of credibility apply for stories written by staff reporters on legacy media websites. The cue that this is a professionally produced story is a more powerful predictor of story credibility than gender. Manipulating theauthor did little to influence experiment participants as far as staff written stories were concerned.

The author matters in professional news, but whether his or her gender matches the audiencemembers does not. This may represent what Armstrong and McAdams (2009) predicted when blogs and other forms of user-generated content become more mainstream. What matters then isthe expertise and professionalism of the author, not whether he is a man or woman. Simplyattaching that author to a respected news organization engenders trust for men and women.

For user-generated content, however, gender stereotypes continue to have some influence, but that influence, for the most part, can be mitigated by how connected a person feels to thestory and the site behind it. Men, for example, found audience written stories nearly as credibleas staff written stories, while they found female staff writers more credible than male reporters. Women, on the other hand, felt more connected with female writers only when they wereaudience members.In other words, gender matters less than social presence, or the measure of how muchreceivers notice the author behind the study. Men may simply chose news that is interesting tothem regardless of who wrote the story. This represents a repudiation of Pearson’s early researchthat male writers are more credible. In fact, this study suggests that men may think that female journalists write more interesting stories.

This study offers some support for the perception that citizen journalism relates more towomen than men. Women are looking for a similar point of view when they approach citizen journalism. They look at other cues when reading staff written stories, but rely somewhat ongender stereotypes for nontraditional stories. This confirms Armstrong and McAdams (2009)finding on blogs and Flanagin and Metzger’s (2003) finding in personal websites. Women aremaking more clear distinctions between citizen journalism and traditional journalism online, andnotice gender distinctions more than men.What they notice most, this study suggests, is the presence of other women in citizen journalism forums and how credible their stories seem, not necessarily expertise or coorientation.

This may support in part the perception that women are more attracted to user-generated content, but not because the stories are produced by other women. They still judge it based on its perceived credibility. In other words, they are not giving content a pass simply because theauthor is another women. The lower scores women assigned to interest may mean they are morecritical consumers of user-generated content even while they appreciate seeing other women getinvolved. The key for them was social presence or noticing the author’s part in the story. Writers, both professional and amateur, who want to reach a female audience, could focus more on distinguishing themselves through their writing, such as being more transparent with their sources of information, the process that went into creating the story, or even their personalreflections on why the story is important.

As Askcherlock, Fowler and Huffington argued, citizen journalism may present new andexpanded opportunities for female voices to join the news process. In fact, this study seems tosupport the idea that women’s voices are vital to the continued growth and success of citizen journalism ventures. As women get involved, however, they must stand on their own merits andnot any roles or stereotypes assigned to them in the past.It is also interesting to note this study found no detrimental effect on organizationalcredibility for publishing audience written content. Audience written stories may have receivedlower story credibility scores, but this difference was largely eliminated when social presenceand interest were added to the equation.

In other words, CNN may not be hurting its brand byinvesting heavily in iReport. In fact, the audience members who write for iReport may help thenews organization connect with and involve more women in the news process, as long as theymaintain their unique voice and make their presence known.This study is limited by the experimental method it used. These findings can only begeneralized for the 175 participants. The study also barely scratched at a complex question withsimple manipulations of authorship and the author’s gender. A more complete study thatattempts to examines which elements predict the concept or credibility, and which help mitigateany gender stereotype effects if any.