Every Blog Has Its Day: Politically-interested Internet Users’ Perceptions of Blog Credibility

Abstract

This study employs an online survey to examine U.S. politically-interested Internet users’ perceptions of the credibility of blogs. The article focuses on the influence of blog reliance compared to motivations for visiting blogs in determining blog credibility. The study found that blogs were judged as moderately credible, but as more credible than any mainstream media or online source. Both reliance and motivations predicted blog credibility after controlling for demographics and political variables. Reliance proved a consistently stronger predictor than blog motivations. Also, information-seeking motives predicted credibility better than entertainment ones.

Introduction

Political pundits have long pondered the importance of television images in presidential elections. But in the 2004 presidential campaign, a new blogger movement arose to challenge the supremacy of television (Bahnisch, 2006; Bichard, 2006; Kerbel & Bloom, 2005; Trammell, 2006). For example, blogs are credited for jump starting Howard Dean’s 2004 presidential campaign (Stromer-Galley & Baker, 2006) and blamed for fueling speculation that John Kerry had won the race based on premature exit poll results (Carlson, 2007).

Bloggers and traditional journalists argue about how much faith to place in messages posted on blogs (Cristol, 2002; Kurtz, 2002). Although blog users are well aware that bloggers are opinionated and attack prominent figures they do not agree with, they still place faith and trust in blog information (Banning & Trammell, 2006; Johnson & Kaye, 2004). It is precisely the perceptions of credibility that give blogs the power to unite followers and bring about social change.

Blog credibility warrants study for several reasons. First, blogs are increasingly becoming an intrinsic part of the Internet. In 1999, the 50 or so existing blogs were largely known only to a few users, but by spring 2006, about 39% (57 million) of all U.S. Internet users accessed blogs (Lenhart & Fox, 2006). Second, scholars suggest that the influence of blogs exceeds their readership. Leading journalists and political officials monitor influential blogs both for information and for story tips (Cassidy, 2007; Singer, 2006). Finally, credibility becomes an important factor when seeking to understand how blogs have so quickly come to challenge online media sites as a leading source of news, especially because people are less likely to pay attention to a medium they do not perceive as credible (Johnson & Kaye, 2000, 2004).

Scholars suggest that credibility is not a characteristic of a source, medium, or message, but is dependent on the perceptions of the receiver (Berlo, Lemert, & Mertz, 1969; Schweiger, 2000). Indeed, studies suggest that credibility of online sources may be linked to the type of sources visited (Flanagin & Metzger, 2000) and to user motivations for accessing such sources (Johnson & Kaye, 2002). However, despite the important role in audience perceptions of credibility, only a few studies have specifically linked credibility with uses and gratifications theory (e.g., Greer, 2003; Johnson & Kaye, 2002; Kim, 2006).

The current study employs an online survey conducted during the 2004 presidential election to determine how believable, fair, accurate, and in depth politically-interested Internet users judge blogs to be. The study also examines whether blog users judge blogs as more credible than do nonusers, and it compares the degree to which blog reliance and blog motivations predict the credibility of blogs, after controlling for demographic and political factors.

Review of Literature

Blog Credibility

Previous researchers have found conflicting results regarding how much the public trusts blogs. Studies have revealed that the degree of reliance on blogs emerges as the key determinant of how much faith is put into them.

When Johnson and Kaye (2004) surveyed blog users during the first month of the Iraq invasion, they found that almost three-quarters of respondents deemed blogs to be moderately to very credible, and only 3.5% rated them as not at all or not very credible. Blog reliance emerged as the only significant predictor of blog credibility. While fewer than four in 10 (38.8%) judged blogs as fair, respondents (59.6%) judged them higher for believability, suggesting that fairness was not the major criterion users relied on when assessing credibility. Rather, they deemed blogs credible because they provided the depth, analysis, and points of view missing from traditional sources.

Johnson and Kaye (2007) discovered that politically-interested Internet users relied more on blogs than on any other news source for news and information and that blogs were judged as more credible than online newspaper sites, online cable television sites, and online broadcast news sites. Similarly, Kim (2006) found that politically-interested Internet users judged blogs as more credible than either portal or mainstream news sites. Blogs scored highest for depth and lowest on fairness, but scores for all credibility measures (accuracy, depth, believability, and fairness) were higher for blogs than for mainstream or portal sites.

In contrast, studies that have examined general Internet users, rather than either blog users or politically-interested users who rely heavily on blogs, have found that blogs rate low on credibility measures. However, scores were higher for those who had visited blogs than those who had not (Banning & Trammell, 2006; Consumer Reports Web Watch, 2005; Hostway, 2005). Additionally, The Consumer Reports Web Watch (2005) found that only 12% of general Internet users believe that information on blogs is accurate, although one reason for the low score is that 31% failed to rate blogs’ accuracy. Among those who had ever visited a blog, 23% said that they trust what they read on blogs at least half of the time.

Therefore, based on previous studies of blog credibility, this study predicts:

H1: Politically-interested Web users will judge blogs to be credible sources of news and information.

H2: Politically-interested Web users will judge blogs to be a) high in credibility for depth of information and b) low in credibility for fairness of information.

This study also examines the following research question:

RQ1: How will politically-interested Internet users judge the credibility of blogs in comparison to traditional and online versions of newspapers, broadcast television news, and cable television news?

Blog Reliance and Credibility

Scholars have debated whether reliance on the Internet in general and on the Web in particular predicts credibility of online information (Greer, 2003; Johnson & Kaye, 2002; Kiousis, 2001). For instance, while Greer (2003) found that Internet reliance proved to be the only predictor of news story credibility, Johnson and Kaye (2002) discovered that Web reliance negatively predicted credibility of online newspaper sources and was unrelated to other sources.

Although findings on the connections between general Internet or Web reliance and credibility are inconsistent, reliance on blogs has emerged as the strongest predictor of blog credibility (Banning & Trammell, 2006; Johnson & Kaye, 2004; Kim, 2006). Users may deem blogs to be credible because they are viewed as being independent from corporate-controlled media (Andrews, 2003; Regan, 2003; Singer, 2006) and, thus, are free to discuss issues more openly and from different perspectives (Bruns, 2006; Cristol, 2002; Wall, 2006). While studies of traditional media suggest that opinionated writing lowers the credibility of the medium (Metzger, Flanagin, Eyal, Lemus, & McCann, 2003), blog users claim that opinionated, thoughtful analysis of the news is precisely what is missing from mainstream media and adds to blog credibility (Bruns, 2006; Lasica, 2002). Experienced, savvy blog users may be more attracted to the personal, opinionated writing style found on blogs than newer users. Further, findings indicate that individuals use various cues, such as reputation and style of delivery, to judge the credibility of a medium (Flanagin & Metzger, 2000; Tewksbury & Althaus, 2000). Inexperienced users may not be as familiar with the overall style and purpose of blogs and thus may view them as less credible.

Based on previous studies of the relationship between blog reliance and credibility, this study predicts:

H3: People who rely heavily on blogs will judge them to be more credible than those who do not, or who rarely rely on, blogs for news and political information.

H4: Reliance on blogs will significantly predict blog credibility, after controlling for demographic and political variables.

Blog Credibility and Political Attitudes

Political variables have been found to be the strongest predictors of blog credibility, with political involvement, political knowledge, political interest, and political trust all predicting blog credibility (Johnson & Kaye, 2004).

Studies suggest that political variables, particularly partisanship and political involvement, may influence credibility judgments because of the hostile media effect. Hostile media effect studies suggest that partisans may judge news coverage to be biased against their own point of view and therefore judge the news content as less credible (Arpan & Raney, 2003; Gunther & Liebhart, 2006; Gunther & Schmitt, 2004), with effects strongest for those who are strongly involved with the issue under study (Arpan & Raney, 2003; Schmitt, Gunther & Liebhart, 2004). Indeed, a recent study found that opponents of the Iraq War perceived that the Internet was less aligned with the government’s pro-war position than traditional media and therefore perceived the Internet as more credible than neutrals or supporters (Choi, Watt & Lynch, 2006).

Blog Credibility and Demographics

Studies indicate that as the Internet becomes more mainstream, demographics have less of an influence on Internet credibility (Johnson & Kaye, 2000, 2002). Political blog users, however, resemble early Internet users, as they tend to be young, well-educated males with high incomes (Eveland & Dylko, 2007; Kaye & Johnson, 2004, 2006; Lenhart & Fox, 2006). Thus it would be expected that in the case of blog credibility, demographics would be influential However, demographics have not emerged as strong predictors of blog credibility. Indeed, both Kaye and Johnson (2004) and Kim (2006) failed to find any demographics that predicted blog credibility.

Gratifications of Blogs

Several researchers have examined what motivates people to seek out blogs (Kaye, 2005, 2007; Kim, 2006; Kim & Johnson, 2005; Seltzer & Mitrook, 2006). Political observers and researchers have identified four main reasons why individuals frequent blogs: community, convenience, to check information found in other media, and information seeking. First, blogs seem to foster a sense of community among users. Blog users can locate blogs that share their viewpoints, as well as discuss their views with the site creator and other users on the site. Second, filter-type blogs are favored because they conveniently aggregate information from many different sites and therefore allow people access to several different perspectives on an issue (Blood, 2003; Kaye, 2007; Papacharissi, 2004; Thompson, 2003; Trammell, Tarkowski, Hofmokl, & Sapp, 2006; Tremayne, Zheng, Lee, & Jeong, 2006). Users may also be motivated to access such blogs for information-seeking purposes, because some blogs are willing to take on issues that corporate-owned media might shy away from and run stories from media outlets throughout the world that were unavailable or just ignored by traditional media (Cristol, 2002). Finally, because of their volume and depth of information, blogs are ideal for instrumental users seeking out information on a specific topic (Singer, 2006).

Indeed, information seeking, convenience, serving as a check on media, and social utility have emerged as major motivations for blog reading (Kaye, 2005, 2007; Kaye & Johnson, 2006; Kim, 2006). Additionally, because blogs are functionally different from both traditional media and their online counterparts, researchers have also found motivations unique to blogging (Kaye, 2005; 2007; Seltzer & Mitrook, 2006). For instance, Kaye (2005; 2007) identified personal fulfillment (a combination of entertainment and social utility motivations), social surveillance (e.g., for a variety of viewpoints, to learn about other’s opinions), and expression/affiliation (to make opinions known, to be in contact with like-minded individuals) as motivations for visiting political blogs.

Gratifications and Credibility

Credibility is a perception held by the audience, not a characteristic inherent in a message, a source, or a media channel (Metzger et al., 2003). Credibility judgments, therefore, are audience centered. Yet few studies have examined how people’s motivations for going online affect their credibility judgments.

People judge various types of online resources differently in terms of credibility because of different expectations about the quality of information that they would find there, which would suggest that different motives for going online have different effects on perceptions of credibility. For instance, information and reference websites have been judged as more credible than entertainment and commercial ones (Flanagin & Metzger, 2000). Also, people judge news sites that are associated with media organizations as more credible than those created by individuals (Cassidy, 2007; Finberg & Stone, 2002).

Further, studies that have examined relationships between motivations and credibility have found that reasons for going online are predictors of Internet credibility. Information seeking motivations, in particular, seem to be more strongly linked to Internet credibility than do entertainment ones (Kaye & Johnson, 2002). This parallels findings from traditional media that suggest that people who consume media for entertainment, diversion, and relaxation needs (e.g., habitual needs) are less goal-oriented and are therefore less selective when choosing media content than those who use media for informational (goal-oriented) needs (Blumler, 1979; McLeod & McDonald, 1985; Rubin, 1984).

Based on previous studies that link motivations to perceptions of credibility, this study predicts:

H5: Motivations for using blogs will significantly predict blog credibility, after controlling for demographic and political variables.

H6: Informational (i.e., instrumental) motivations will be a stronger predictor of blog credibility than will entertainment motivations (i.e., habitual), after controlling for demographics and political variables.

Gratifications vs. Reliance

Only a few studies have directly compared online motivations with Internet reliance to determine which is the best predictor of credibility. These studies have produced different results. Johnson and Kaye (2002) discovered that traditional media reliance was the strongest predictor of credibility of online media and that convenience as a motivation was the second strongest predictor. Greer (2003) found that Web reliance was the strongest factor explaining why an online news story was judged as credible, while the four motivation measures (information seeking, passing time, communication, and social) did not impact credibility judgments. Finally, Kim (2006) discovered that while blog reliance proved the strongest predictor of credibility judgments, using blogs for convenience/information seeking and entertainment also significantly impacted credibility judgments.

Previous studies that compared the influence of motivations and reliance on perceptions of credibility have differed on how strongly they found that motivations explained credibility, but all indicated that reliance was a stronger predictor. Therefore, this study predicts:

H7: Blog reliance will prove a stronger predictor of blog credibility than will motivations, after controlling for demographic and political variables.

Method

Politically-interested Internet users’ judgments of credibility of blogs were gathered from a survey that was posted online from October 19 to November 16, 2004—the four-week period surrounding the 2004 presidential election day. A brief description of the survey, along with a general statement of purpose, author biographies, and survey URL were sent to media-focused and politically-focused websites,¹ electronic mailing lists and bulletin boards,² and blogs³ that were most likely to attract politically-interested users across a range of political ideologies. Blog postings usually included a sentence of support from the blogger encouraging users to fill out the survey. The study also employed a snowball sample, as individuals were asked to forward the survey to politically-interested friends. This convenience method of obtaining respondents yielded 1,399 completed surveys.

Motivations for Using Blogs

The motivations for using blogs during the 2004 presidential election were adapted from Kaye’s (2005) study of gratifications for using blogs during the Iraq War. Kaye identified six motivations for accessing blogs: Information Seeking/Media Checking, Convenience, Personal Fulfillment, Political Surveillance, Social Surveillance, and Expression/Affiliation. Only Expression/Affiliation did not achieve a Cronbach alpha reliability score of .70 or higher.

For the current study, additive indices were created for the five significant motivations: Information Seeking/Media Checking—actively searching out information about current issues, analyzing blog information, and comparing accounts with those found in traditional media; Convenience—using blogs because it is convenient and easier to do than turning to traditional media or other online sources, such as bulletin boards or chat rooms; Personal Fulfillment—fulfilling various emotional needs and gathering information to use in discussions with others; Political Surveillance—keeping an eye on the political landscape; and Social Surveillance—learning about others’ points of view and opinions on various issues and current events. Cronbach alphas were run on each of the five sets of motivations to check for internal reliabilities, with scores ranging from .73 for Social Surveillance to .87 for Information Seeking/Media Checking.⁴

Reliance on Blogs

Reliance on blogs was measured by a single question that asked respondents whether they (5) “heavily rely,” (4) “rely,” (3) “sometimes rely,” (2) “rarely rely,” or (1) “never rely” on blogs for political information. Next, respondents were categorized according to their level of reliance on blogs, to compare credibility differences between high and low users. Those who “heavily rely” or “rely” on blogs were grouped as “heavy reliance” respondents. Those who “rarely rely” or “never rely” on blogs for political information were combined as “light reliance” users. Respondents who marked that they “sometimes” rely on blogs were excluded from further analysis, in order to provide a clearer distinction between high reliance and low reliance.

Credibility of Blogs

Using past studies as a guide (Gaziano & McGrath, 1986; Johnson & Kaye, 2000, 2002, 2004; Meyer, 1988; Newhagen & Nass, 1989), credibility was measured as a multidimensional construct. Believability, fairness, accuracy, and depth of information are four measures that have consistently emerged as variables that best gauge media credibility (Gaziano & McGrath, 1986; Meyer, 1988; Newhagen & Nass, 1989) and thus were used in this study.

Respondents were asked how believable, fair, accurate, and in depth they judge blogs, online and traditionally delivered newspapers, traditional and online television news programs, and traditional and online cable news shows, using a five-point scale, with 1 indicating “not at all” and 5 indicating “very.” The four measures were then combined into a credibility index. Cronbach’s alpha for the measures are as follows: blogs .87, online broadcast television .81, broadcast television .89, online cable news .89, cable news .91, online newspapers, .93, and online newspapers .93.

Political Characteristics

Levels of political involvement, political knowledge, political interest, interest in the 2004 presidential election, trust in the government, and self-efficacy were assessed using various scales. Degree of political involvement was determined by asking respondents whether their involvement in politics since becoming an Internet user has (5) “greatly increased,” (4) “increased,” (3) “stayed the same,” (2) “decreased,” or (1) “greatly decreased.” Respondents were asked how knowledgeable they are about politics in general, how interested they are in politics, and how interested they are in the 2004 presidential campaign. These variables were measured on a 10-point scale, with 0 indicating not at all knowledgeable or interested, and 10 signifying highly knowledgeable or interested. This semantic differential scale was used as a variation on the Likert scale to obtain interval data for statistical analysis.

Self-efficacy and trust were both measured using a five-point scale: (5) “strongly agree,” (4) “agree,” (3) “neutral,” (2) “disagree,” or (1) “strongly disagree” with the self-efficacy and trust items. The self-efficacy and trust items were the same as those used in the National Election Studies. Self-efficacy was assessed by the following three statements: “I consider myself well qualified to participate in politics,” “I think that I am better informed about politics and government than most people,” and “I feel that I have a pretty good understanding of the important political issues facing our country.” Trust in the government was measured by “Most of our leaders are devoted to the service of our country,” “Politicians never tell us what they really think,” and “I don’t think public officials care much about what people like me think.” The polarity was reversed on the last two statements of the trust index. The trust and efficacy measures were combined into separate self-efficacy and trust indexes (3-15 score range). The reliability for the self-efficacy and trust indexes is .73 and.75, respectively.

Demographics

The four demographic measures were gender, age, education, and income. Respondents marked their gender and were asked their age as of their last birthday. They also selected the highest grade completed from the following list: “less than high school,” “high school graduate,” “some college,” “four year college degree,” “master’s degree,” “Ph.D. degree,” and “other,” and they estimated their 2004 income.

Data Analysis

First, mean scores were calculated for each of the individual measures (believable, fair, accurate, depth) of credibility of blogs and the other traditional and online media. Next, believability, fairness, accuracy, and depth were combined into one index of credibility, with scores ranging from 4 to 20. Then paired sample t-tests compared judgments of credibility of blogs to each of the six sources (traditional and online newspapers, traditional and online television, traditional and online cable news) under study. Next, independent sample t-tests compared the differences in perceptions of blog credibility between respondents who heavily rely on blogs to those who lightly rely on them.

Last, the motivational items were combined into five factors and reliability scores calculated. The factors were then used as independent measures, along with reliance, political characteristics, and demographics, in hierarchical regression. Regression analysis was used to determine the predictive power of reliance on blogs and motivations for using blogs on perceptions of credibility after controlling for the other variables. Demographics were entered as the first block. The second block consisted of the political variables (political involvement, political knowledge, political interest, election interest, self-efficacy, and trust). Reliance was entered into the last block with one of the five motivation factors. The variables were entered in ordered blocks to control for influence on reliance and motivation. Regressions were repeated five times, each with a different motivation factor.

Results

Respondent Profile

More males (62.6%) than females (37.4%) completed the survey. Nearly all (97.3%) were educated, with at least some college, and the average age was 43.4 years. More than nine in 10 (91.2%) of the respondents were white, 1.3% black, and the remaining classified themselves as other non-whites. The average annual income was just over $64,000. The demographic profile of this study closely matches the characteristics of respondents in studies that have examined blog users; people who seek out information from blogs tend to be well-educated, white middle-class males (Consumer Reports Web Watch, 2005).

The overwhelming majority (92.4%) of respondents expressed a strong interest in the 2004 presidential campaign, and just under three-quarters (73.6%) agreed that they were very interested in politics in general. Further, 87.6% reported that they were very knowledgeable about the 2004 election, and almost two-thirds (64.2%) were highly knowledgeable about politics in general.

Four out of 10 (41.1%) respondents moderately trusted the government, with only 28.4% reporting high levels of trust. Despite low faith in the government, almost all (97.2%) of the respondents were confident in their power to bring about change and therefore reported high levels of self-efficacy.

Almost half of the respondents (48.6%) claimed that they were strongly affiliated with their party of choice, whether Republican (30.5%), Democrat (35.4%), Libertarian (5.4%), Green (1.2%), other party (1.6%), or independent (15.6%). Further, 38.1% consider themselves conservative/very conservative, 25.6% are moderate, and the remaining 36.3% are liberal/very liberal.

When seeking political information, these respondents rely heavily/rely on blogs (63.4%), followed by political websites (56.7%), mailing lists/bulletin boards (35.1%), Web portals (30.0%), and chat rooms (8.5%). Further, the respondents spend an average of 8.3 hours per week on political sites, followed by 7.4 hours per week on blogs.

Credibility of Blogs and Traditional Media

The first hypothesis predicts that politically-interested Internet users would judge blogs as credible sources. While mean scores (range 4–20) for blogs are higher (12.2) than for any other source overall, respondents judge blogs as only moderately credible. Hypothesis 1, then, is partly supported (Table 1).

Previous studies (Bruns, 2006; Cristol, 2002; Wall, 2006) suggested that because bloggers voice their opinions on issues of the day, blogs are not rated highly for fairness. Of the four measures of credibility, blogs are rated the lowest for fairness (M = 2.6), as hypothesized. However, blogs are deemed more fair-minded than the other sources under study. It was also hypothesized that because blogs feature information missing from other sources and they discuss issues in more depth than do mainstream media, users would rank blogs high for depth of information. Indeed, that was the case (m = 3.58). Hypothesis 2 is supported; respondents deem blogs low on fairness but high in depth of information. It is worth noting that among all media, traditional broadcast television news suffers the lowest credibility in all aspects (believability m = 2.35, fairness m = 2.05, accuracy m = 2.26, depth m = 1.66) (Table 1).

Credibility of Online and Traditional Media

The research question asks whether politically-interested Internet users judge blogs as more or less credible than traditional media or other online sources. Blogs were ranked the most credible (M = 12.2) of all media: traditional newspapers (M = 11.2), online newspapers (M = 11.2), traditional television news (M = 8.3), online television (M = 9.2), traditional cable TV news (M = 9.9), and online cable television news (M = 10.5). Paired sample t-tests were used to compare the credibility of blogs with other media sources. The perceived credibility of blogs is significantly higher than the perceived credibility of any other traditional and online media (Table 2).

Perceptions of Credibility: Heavy vs. Light Blog Users

The third hypothesis tests the different perceptions of credibility between heavy blog users (n = 861) and light users (n = 285). Independent sample t-tests indicate that heavy blog users judged blogs as significantly more credible (believable, fair, accurate, in-depth, combined) than did light users (t = −27.9; df = 1068; p = .000). Hypothesis 3, then, is supported. Furthermore, heavy users also ranked blogs as higher on all of the individual measures: believability (t = −24.9; df = 1083; p = .000), fairness (t = −16.3; df = 1081; p = .000), accuracy (t = −23.8; df = 1078; p = .000), and depth of information (t = −23.7; df = 1075; p = .000). Of the four attributes, blogs are judged higher in credibility for depth of information than other media by both heavy (M = 4.1) and light users (M = 2.3) and lowest in credibility for fairness of information (M = 2.9 and 1.7, respectively), as Table 3 indicates.

Predictors of Blog Credibility

Reliance and Motivation

This study predicts that both reliance on blogs and motivations for using blogs would be significant predictors of blog credibility, after controlling for political variables and demographics. Indeed, both significantly predict credibility. The combination of reliance on blogs and motivations explain between 41% and 45.7% of the perceptions of blog credibility. Hypotheses 4 and 5 are supported (Table 4).

Reliance on blogs is positively associated with judgments of blog credibility, as are the motivation factors. Among the five motivations, political information seeking is the strongest predictor (b = .36, p < .001) and personal fulfillment/entertainment is the weakest predictor (b = .13, p < .001). Furthermore, three informational motivation factors—information seeking (b = .36, p < .001), political surveillance (b = .31, p < .001), and social surveillance (b = .26, p < .001)—are stronger predictors than the entertainment factor and personal fulfillment (b = .13, p < .001). Thus, instrumental motivations impact blog credibility more significantly than do entertainment motivations, after controlling for demographics and political variables. Therefore, hypothesis 6 is supported (Table 4).

The final hypothesis asserts that blog reliance will be a stronger predictor of credibility than will motivation, after controlling for demographic and political variables. Reliance on blogs was found to be a positive predictor of credibility, and it is a stronger predictor than any of the motivational factors, after controlling for demographic and political variables (Table 4). Therefore, hypothesis 7 is supported.

Demographic Variables

Gender and income consistently predict the assessment of blog credibility. Men consistently judge blogs as more credible than do women. Users with low incomes are more likely to rate blogs as credible than are users with high incomes. Demographic variables account for an additional 13% of the variance in the perception of blog credibility (see Table 4).

Political Variables

Political variables explain 56.4% of the variance in the assessment of blog credibility. Among all six political variables, political involvement, political interest, and election interest consistently predict blog credibility. Both election interest and political involvement are significant and positive predictors for the perception of blog credibility. However, political interest significantly, but negatively, impacts the assessment of credibility. Those whose involvement in politics has greatly increased since going online rated blogs as credible sources of information. In contrast, the more interested respondents are in politics in general, the less likely they are to judge blogs as credible. Neither trust nor efficacy is a significant predictor of blog credibility (see Table 4).

Discussion

This investigation employed an online survey of politically-interested Internet users during the 2004 presidential election in an effort to examine perceptions of blog credibility. The study addressed judgments regarding the believability, fairness, accuracy, and depth of information provided in blogs. Differences in assessments of blog credibility were evaluated for both heavy and light blog users. Reliance and motivations for blog use were also explored as possible predictors of blog credibility. Despite the important role of audience motivations for visiting a site in their perceptions of credibility, only a few studies have specifically linked credibility with uses and gratifications theory (e.g., Greer, 2003; Johnson & Kaye, 2002; Kim, 2006). This is an important connection, because motives for going online can determine perceptions of credibility. To complete the analysis, blogs were then contrasted with other media outlets on judgments of credibility.

This study hypothesized that politically-interested Web users will judge blogs as credible sources of news and information. The results indicate that politically-interested Internet users find blogs to be moderately credible sources for news and information. Past analyses have detected similar findings for blogs (Banning & Trammell, 2006; Johnson & Kaye, 2004) and for the Internet as a whole (Johnson & Kaye, 2000, 2002). While Internet users are increasingly flocking to blogs as a source of political news and information, the moderate scores for credibility indicate that users also realize that blogs are not the final word.

As expected, respondents evaluated blogs as highly credible for depth of information, while judging them as weaker on credibility for fairness of information. Johnson and Kaye (2004) assert that fairness is not the primary concern for users when judging blog credibility. The researchers suggest that the depth of information presented is a more important indicator than fairness in assessing credibility. Users may rely on blogs to search for information that appears to be missing from more traditional media coverage or may be covered too lightly. Blogs allow for the presentation of up-to-date, often candid viewpoints of specific interest for their users. Past research indicates that while traditional media are expected to maintain standards of fairness and balance (Metzger et al, 2003), such expectations are not extended into the blogosphere (Bruns, 2006; Lasica, 2002). In fact, blog supporters perceive bias as a strength that allows for a more detailed and in depth examination of issues. Although this intensity of analysis is valued, it may also suggest that blog readers are attracted to sites that share their viewpoints and therefore have a potentially polarizing effect.

Heavy users viewed blogs as more credible overall than did light blog users. This perception extended to all four measures of credibility (believability, fairness, accuracy, depth of information). Traditional media and their online forums provide cues of authenticity and reputation (Flanagin & Metzger, 2000; Tewksbury & Althaus, 2000) that may influence judgments of credibility. Such cues may be missing from blogs. Inexperienced users, who are unfamiliar with the often opinionated and non-traditional format of blogs, may find them a less credible source of information. It takes more experience to truly engage with blogs, from navigating the Web to finding political sites with topics of shared interest. Additionally, more experienced blog users may be more adept at searching the blogosphere for political opinions similar to their own (Johnson & Kaye, 2004) and may therefore judge blogs as more credible. These results support previous studies that suggest that the more a user relies on a source, the more likely he/she is to view it as credible (Wanta & Hu, 1994; Westley & Severin, 1964).

Politically-interested blog users judged blogs as considerably more credible overall than traditional media or other online sources. Interestingly, broadcast television fared the worst on measures of credibility. The hostile media effect may play some role in this perception of credibility. If blog users perceive mainstream news coverage as biased against their point of view (Arpan & Raney, 2003; Gunther & Liebhart, 2006), they may be more likely to deem it not credible. For example, past research indicates that conservatives are more likely to rely on blogs (Johnson, Kaye & Kim; 2006), because they view the mainstream media as dominated by a liberal viewpoint (Copeland, 2004; Hamdy & Mobarak, 2004). Perhaps blog content is perceived as more credible because it is considered more independent than news reported in corporate-controlled media. Also, blogs provide an open forum for users to address issues (Bruns, 2006; Cristol, 2002; Wall, 2006). Users may also gravitate toward blog content because it allows for a more personal and interactive venue for acquiring information (Papacharissi, 2004; Reynolds, 2005; Thompson, 2003; Wall, 2005) than traditional media. It is therefore vital to understand blog users’ motivations for usage in order to gain further knowledge about factors that influence the degree to which they are viewed as credible.

As hypothesized, both reliance and motivations were predictors of blog credibility, after controlling for demographic and political variables. Reliance emerged as a stronger predictor of blog credibility than did motivations. It stands to reason that users would find their preferred media selection the most credible. Past analyses have also indicated that reliance is a consistently strong predictor of credibility perceptions (Banning & Trammell, 2006; Johnson & Kaye, 2004; Kim, 2006). All five of the motivation factors under scrutiny were also positive predictors for blog credibility (Information Seeking/Media Checking, Convenience, Personal Fulfillment, Political Surveillance, and Social Surveillance). These findings support other studies that have asserted that bloggers are motivated by a variety of reasons for going online and that different sites cater to varying user desires (Flanagin & Metzger, 2000; Johnson & Kaye, 2004).

The current study took the examination one step further by asserting that the Informational (i.e., instrumental) motivations will be a stronger predictor of blog credibility than will entertainment motivations (i.e., habitual), after controlling for demographics and political variables. When assessing the influence of specific motivations for blog users, political information seeking emerged as the strongest predictor for judgments of blog credibility, while personal fulfillment/entertainment was the weakest. Past analyses indicate that political surveillance is a strong motivator for blog users (Kaye, 2005). The Internet is consistently viewed as a resource for more instrumental needs such as surveillance, voter guidance, and social utility (Lin & Salwen, 2006; Lin, Salwen, & Abdulla, 2005) Perhaps individuals are attracted by blogs because they offer a wide variety of issue perspectives, as well as opportunities for political discussions.

However, respondents who primarily engaged in blogging for entertainment purposes are less likely to judge blogs as highly credible. It stands to reason that credibility factors such as believability, fairness, accuracy, and depth may not be as important for those engaging in blogging for fun and personal fulfillment. Studies suggest that those who consume media for entertainment, diversion, and relaxation needs are less goal-oriented and are therefore less selective when choosing media content (Blumler, 1979; McLeod & McDonald, 1985; Rubin, 1984). The findings are encouraging, in that people seriously motivated by political examination indeed find blogs to be credible sources for their information inquiry.

Limitations and Suggestions for Future Study

The respondents in this study were a self-selected group of politically-interested Internet users. While a random sampling would have been ideal, potential respondents were reached through announcements posted on politically-oriented websites, bulletin boards/lists, blogs, and chat rooms. Purposive sampling techniques, such as that used in this study, are often used to generate results that are representative of a subset of users, but are not necessarily representative of the larger population (Babbie, 1990). The Internet does not offer tools for random sampling of the entire Internet population or for blog users in particular; therefore non-probability sampling is an appropriate technique for posting an online survey.

Overall, the current findings highlight various perceptions and predictors of blog credibility among politically-interested web users. Judgments of credibility were highest for information depth and were largely predicted by degree of reliance and motivations. The analysis provides deeper insight into the distinct motivations of blog users and their resulting opinions of blog content. Future research is needed that compares the current findings to future elections and individuals as they engage with new media of all types. Issues of polarization should be addressed as well, to discover if individuals simply rely on new media sites that match their current viewpoints or seek out different perspectives. This study, in line with earlier studies, treated credibility as a dependent variable. However, while this study looked at how reliance predicted credibility, it is also true that credibility may determine how much people rely on the media. Future studies may want to explore the causal relationships between reliance and credibility further.

Future study may also benefit from an examination of the differences in motivations for blog users and the impact on perceptions of blog credibility. Specific uses and gratifications would appear to have a substantial impact on those perceptions. The ability to better understand and predict perceptions of credibility would be a valuable resource for both traditional journalists and bloggers alike. Finally, this study was conducted during the 2004 election, when blogs were gaining popularity. As more people are relying on blogs for news and information, credibility scores may have increased, and reliance and motivations may prove to be even more strongly linked to blog credibility. This study should be replicated in the 2008 presidential election to determine whether people judge blogs to be more credible and whether factors influencing blog credibility have changed.

Notes

References

About the Authors

Thomas J. Johnson [[email protected]] is the Marshall and Sharleen Formby Regents Professor in the College of Mass Communications at Texas Tech University. His research interests lie in the areas of new media technology, credibility, and political communication.

Address: College of Mass Communications, Texas Tech University, Box 43082, Lubbock, TX 79409, USA

Barbara K. Kaye [[email protected]] is an associate professor in the School of Journalism & Electronic Media at the University of Tennessee—Knoxville. Her research interests include media effects and consumer uses of new communication technologies. More specifically, she researches how Internet and blog use influence political attitudes and media use behavior. She also studies the uses and effects of profanity on broadcast and cable television programs.

Address: School of Journalism & Electronic Media, University of Tennessee, 333 Communications Building, Knoxville, TN 37996, USA

Shannon L. Bichard [[email protected]] is an assistant professor in the College of Mass Communications at Texas Tech University. Her research interests include the areas of political communication, public opinion, and consumer behavior.

Address: College of Mass Communications, Texas Tech University, Box 43082, Lubbock, TX 79409, USA

W. Joann Wong [[email protected]] is an assistant professor in the School of Journalism at Indiana University. Her research interests focus on political and international communication.

Address: School of Journalism, Indiana University, 940 E. Seventh St., Bloomington, IN 47405, USA