Abstract
Photojournalists publish images they have created in news publications and on social media, and images captured by ordinary citizens sometimes appear in journalism spaces. This study examines how the professionalism of a photograph’s authorship and presentational context influence the perceived credibility of the image using a two (photographer; staff or amateur) by two (image presentational context; news site or social media) quasi experiment. The small difference in how respondents rate the credibility of the images suggests that, broadly, participants in this study are willing to accept newsworthy images as credible on social media, and social media images as credible in the news.
Introduction
The front page of the Los Angeles Times on March 5, 2022, featured a striking image taken by staff photographer Marcus Yam. In the photo, more than a dozen Ukrainians push onto a crowded train in the town of Irpin as a lone woman hurries, suitcase in hand, to find another open door (Appendix A). Smoke from heavy fighting during the Russian invasion hazes the sky in the background. The L.A. Times’s Photo Department posted this image taken by Yam on Instagram for its approximately 222,000 followers. Users posted comments like “heartbreaking,” and liked the image more than 7,000 times. When a user in Japan asked, “Can I share your photo on instagram?” the Pulitzer Prize winning photojournalist replied, “Thank you for sharing!” (Appendix B). Thus, a prominent image taken by a professional photojournalist in Ukraine, paid for by a California-based news organization, and posted on a commercial social media platform was ultimately shared by an amateur Instagram photographer whose previous two posts featured the first budding trees of Japanese springtime.
This exchange reflects a common occurrence: a newsworthy photo travels quickly across the internet, contextualized by the online platform in which it appears and the source authorship of the image. In this example, a professional journalistic photo traveled from a traditional news medium to an official Instagram account, and from there dispersed more widely via an nonprofessional sharer. Opposite scenarios occur, too, such as when an amateur photographer posted an image of the emergency landing of US Airways Flight 1549 in the Hudson River to social media, and that image was published extensively in professional news contexts (Associated Press, 2009). The modern cascade of imagery includes both professional and amateur-authored visuals, and viewers may encounter either category of news image on a professional news website or on a social media platform. Ultimately, a photograph’s source, authorship and presentational context can quickly become less clear to audiences (Sundar & Nass, 2001). And while studies have examined the impact of source and author on sharing intentions and credibility perceptions, few have considered how audiences evaluate the intertwinedness of presentational and authorship variables specifically pertaining to credibility perceptions of news visuals. With the phenomenon of traditional news appearing on social media decreasing credibility perceptions of the industry as a whole (Karlsen & Aalberg, 2021; Sterrett et al., 2019) and with over 48% of Americans getting their news from social media at least sometimes (Walker & Matsa, 2021)—the question of how news audiences are affected by the combined variables of source image authorship and final presentational context is relevant to journalistic practice.
The purpose of this study is to examine how the credibility perceptions of a newsworthy image might be affected based on the context of its presentation and whether the image was authored by a professional or amateur photographer. The study also considers the interaction and interplay of permutations using these two variables. Using a survey on Amazon MTurk (n = 4,486), credibility perceptions were measured in four presentational contexts: (1) professional photographs on professional news websites (2) amateur photos on professional news websites, (3) amateur photos on social media, and (4) professional photographs on social media. Survey respondents were randomly assigned one of these four variable stimuli for three separate news stories. For each story, survey respondents saw either an image taken by a professional photojournalist or an image taken by an amateur photographer of the same news event. The three news events represented were a protest over the death of Breonna Taylor in Louisville, Kentucky, the January 6 pro-Trump insurrection at the United States Capitol, and the California “Dixie” wildfire in 2021. These high profile events were chosen as stories that gained attention in both traditional news outlets and social media conversations.
The source, medium, and presentational context of news has been shown to affect credibility perceptions and the subjective judgment by viewers about trustworthiness in traditional media (Tseng & Fogg, 1999). Medium credibility refers to the perceived level of credibility a specific medium that individuals used (Newhagen & Nass, 1989; Sundar & Nass, 2001), while source credibility rates the authors who provided the information, focusing on the expertise or trustworthiness of the source and the likelihood they will provide credible information (Armstrong & Nelson, 2005; Flanagin & Metzger, 2003; Hovland et al., 1953). Thus, misunderstood photo sources and medium-of-origin will likely affect perceptions of the credibility of the photos.
A better understanding of the interplay of factors that make an audience perceive an image as more or less credible could ultimately help news organizations better optimize their role in the important task of showing the public what is true.
Literature Review
The literature that guides this study first includes a brief overview of the concept of credibility, defining the broad term and specifically addressing the types of credibility pertinent to the present study: medium credibility and source credibility. The literature of medium credibility perceptions that consider photographs will be included in this section. Studies that have addressed source credibility as it pertains to nonprofessional versus professional content will then be reviewed, and again, the thin literature specific to photography will be included. Finally, an overview of researched differences between professional and nonprofessional photography will be given, providing insight into some of the variables that might sway viewer credibility perceptions.
Credibility
The concept of credibility broadly revolves around an audience’s perceptions of the believability of information (Tseng & Fogg, 1999). Research of media credibility began with a 1936 study by Mitchell V. Charnley on newspaper accuracy. In another seminal study, Hovland and Weiss (1951) named two basic components of credibility: trustworthiness and expertise. These early-identified components were added to, refined, and parsed into more specific variables in the following decades of credibility research to include variables such as fairness, completeness, reliability, motivation for money, respect of privacy, community well-being, objectivity, competence, and trustworthiness, and others (e.g., Gaziano & McGrath, 1986; Kiousis, 2001; e.g., Metzger et al., 2003; Meyer, 1988; Rimmer & Weaver, 1987). These credibility variables measure an audience’s subjective judgment about the trustworthiness of a medium, source, or message (Self, 1996). Medium credibility has traditionally referred to the perceived credibility of a specific medium that people have used, and source credibility focuses on the perceived expertise or trustworthiness of the source of that information.
Professional journalists are tasked with maintaining the credibility of their own work and the news industry as a whole through a strong and historic commitment to truth-telling (Kovach & Rosenstiel, 2021). The Society of Professional Journalists (SPJ) Code of Ethics warns journalists to avoid actions that “may damage credibility,” and the NPPA Code of Ethics implores practitioners “to strengthen public confidence in the profession” through the standards and ideals laid out in the code (SPJ Code of Ethics, 2023). Trust toward the news impacts the manner in which people engage with, pay attention to, and share the news (Sterrett et al., 2019; Tsfati, 2003) Untruthful news can be harmful to viewers who rely on it to make informed decisions and can ultimately damage the credibility of the industry (Elliott, 2011; Silva & Eldridge, 2020; Self, 1996). This commitment to ethics is what distinguishes journalism from other content gathering and disseminating activities (Singer, 2007). Journalists are also meant to be objective, though this term is frequently misunderstood in the popular imagination, as journalistic objectivity is not about the personality of the journalist but rather “a disciplined unity of method transparently conveyed” (Kovach & Rosenstiel, 2021). Credibility is also a core principle in the professional practice of photojournalism. Photographs taken by photojournalists are meant to be taken by practitioners committed to ethical values and deep-set norms (Elliott, 2011; Fahmy et al., 2014). The National Press Photographers Association says that “the credibility of our profession is the responsibility of every member of the NPPA” (Code of Ethics for Visual Journalists, n.d.).
Despite these efforts, social media can “mask or complicate the underlying source of information” when it also appears on a number of different media (Schmierbach & Oeldorf-Hirsch, 2012). In an early recognition of the problem of medium and source information confusion, the authors illustrate, “A person might find the story on the organization’s Web site, find a link from a prominent blog, be sent an email recommending the story or the blog post, or come upon a tweet referencing the story. In such circumstances, opinions about the initial news outlet may become intertwined with opinions about blogs or Twitter, as well as the individuals making the recommendation” (Schmierbach & Oeldorf-Hirsch, 2012, p. 318). In this study, we are interested in the effects of both medium and source as well as their interaction upon credibility perceptions. What proceeds is a review of previous studies of perceptions of a medium’s credibility, including the very little research that has focused on media visuals.
Medium Credibility Perceptions
Medium credibility refers to the perceived credibility of a specific medium that people have used. Mid-to-late century studies typically focused on perceptions of different traditional media, such as newspapers versus television. In one of the most-cited works, Gaziano and McGrath (1986) designed a 12-item scale resulting from a survey to the American Association of Newspaper Editors (ASNE). The scale items concerned whether the media are fair, unbiased, tell the whole story, are accurate, respect people’s privacy, watch out after people’s interests, are concerned about the community’s well-being, separate fact and opinion, can be trusted, are concerned about the public interest, are factual, and have well-trained reporters. The Gaziano and McGrath scale has been used, adapted, and altered by a number of other scholars, including Rimmer and Weaver (1987), Newhagen and Nass (1989), Meyer (1988), and others. Studies in the late 1990s and early 2000s frequently turned to perception studies of the credibility of online media versus other media (e.g., Flanagin & Metzger, 2000; Johnson & Kaye, 1998; Kiousis, 2001; Nah & Chung, 2012; Winter & Krämer, 2014). Much of this prior research has shown that online information is considered about as credible or even more credible than offline information. Johnson et al. (2007) found that some online news sources were rated more credible than their offline counterparts. Kiousis (2001) found that online news was perceived as more credible than television, but less credible than newspapers, and Flanagin and Metzger (2000) reached similar findings.
In many cases, information presented on traditional news mediums–online or off–is perceived as more credible than when presented on social media (Agadjanian et al., 2022; Karlsen & Aalberg, 2021; Schmierbach & Oeldorf-Hirsch, 2012; Schultz et al., 2011; Wada, 2018). Readers generally perceive social media shared by a trusted personal contact rather than the original news organization to be more credible (Cooley & Parks-Yancy, 2019; Sundar & Nass, 2001). Furthermore, people tend to trust posts with a greater number of positive and supportive comments (Mena et al., 2020; Messing & Westwood, 2012; Sundar, 2008). The interaction of visual source and medium with regard to social media and traditional media outlets has not been fully studied, and research regarding credibility perceptions of news from different sources and media has nearly excluded visuals entirely. As Sterrett et al. (2019) point out, “In a digital news environment, traditional conceptions of trust that rely on clear distinctions between source, message, and medium may be inadequate” (p. 786).
Issues of social media credibility have been the subject of research in some of the most recent studies and have broadly shown less trust toward news that is presented on social media. Schultz et al. (2011) were interested in credibility perceptions of information presented by three media: a newspaper, a blog, or Twitter. In general, the results showed that individuals receiving a message via Twitter had a lower impression of the company who sourced the information. Similarly, Schmierbach and Oeldorf-Hirsch (2012) found that Twitter was considered less credible than various forms of stories posted on a newspaper website. Most respondents in a later study by Wada (2018) admitted to rarely checking the source of their information retrieved from social media, but expressed doubts about its credibility. In an experiment by Karlsen and Aalberg (2021) one group of participants was exposed to a news story on an original news website, and another group was exposed to the same story, but presented on Facebook. Consistent with earlier studies, participants were generally less trustful of the news they consumed through social media. The authors emphasize that traditional news appearing on social media can ultimately decrease trust in the news industry.
Sterrett and colleagues (2019) were interested in the impact on credibility of both a sharer of news on social media as well as the original source. Their findings showed that the sharer greatly impacted both trust and engagement with news on social media. But the reporting source significantly affected only engagement. The sharer had a much larger and more consistent impact than the source on both trust and engagement. The sharer and the source did not interact to impact trust or engagement. In another recent experimental study of medium credibility perceptions of traditional versus social media, one group of participants was shown a news article as it would appear on an online news website and a second group was shown Facebook previews of the same news article as if it had been posted by the news organization’s account. Participants exposed to the Facebook condition rated the information as less credible than on news websites, but the effect was small (Agadjanian et al., 2022).
The aesthetics of visual presentation and context also can have significant downstream effects on how the audience perceives the overall credibility of an online news source. For example, factors like not having any images, having too many images or even making a font too large negatively affects the perceived credibility of a news presentation (Wobbrock et al., 2019).
Medium Credibility Perceptions of Visuals on Social Media Versus Traditional Media
Social media is a personalized experience, and research shows how a variety of personal and external factors might influence how viewers judge the credibility of an image on a feed. Factors like a viewer’s internet skills, social media usage and photo editing experience might affect how credible a viewer finds an image presented in a news context (Kasra et al., 2016; Shen et al., 2018). The emotion on the faces of image subjects, as well as the overall positivity of a user’s social media presence, also has bearing on how credible viewers judge those images to be (Djafarova & Rushworth, 2017; Karduni et al., 2021).
Much of the research on credibility perceptions of visuals on social media has been within the realm of marketing. This research has explored what factors make an Instagram profile most credible from a commercial perspective and find clear patterns in what factors an audience uses to decide. To cite just two examples, celebrities are considered more credible on Instagram than non-celebrities (Djafarova & Rushworth, 2017), and trust in branded posts by influencers increases when the influencer is attractive (Lou & Yuan, 2019). Unsurprisingly, less research exists studying credibility perceptions for news imagery on social media.
Source Credibility
Source credibility focuses on the perceived expertise or trustworthiness of the content’s author to provide credible information. Source credibility was a more straightforward concept when journalists had clear bylines and broadcast reporters had lower thirds (Tandoc, 2019). However, in the online environment, people assess the credibility of information using various source cues. Beyond the information’s original source and context, there also exists what Sundar and Nass (2001) refer to as the “selecting source” and what Sterrett et al. (2019) term the “sharer.” This describes the individual or organization that suggests or shares a piece of information but did not gather the information originally. The Sterrett et al. (2019) study found that the overall trust perceived in the sharer of information was more important than overall trust of the source where the information originated. In the Sundar and Nass study, participants perceived information that was said to have been selected by other users more credible than information that was said to have been selected by expert editors. Their findings are generally consistent in follow-up studies.
Overall, social media users tend to trust those they know personally on social media more than individuals they do not (Cooley & Parks-Yancy, 2019). The so-called “bandwagon heuristic” (Sundar, 2008) is related to the phenomenon of trusting personal contacts. It refers to the phenomenon that the more likes and positive or supportive comments a particular post has, the more likely a user will find the post credible (Mena et al., 2020). Research focused on news content specifically shows that the bandwagon heuristic applies, with more recommendations from trusted social contacts equaling more engagement with a particular news story (Messing & Westwood, 2012).
Tandoc (2019) found more subtle results. In examining whether perceived credibility varies when it is shared by a traditional news organization on Facebook or by a friend on Facebook, participants stated that they view their friends as more credible. In the experiment, however, participants rated news articles as more credible when they are shared by a news organization, but only in high-motivation conditions. Impressions about the accuracy of news on social media has been shown to have little effect on sharing intentions, however, as viewers are impacted by other factors such as political concordance, despite proclaiming the importance of only sharing accurate news (Pennycook et al., 2021).
Source Credibility of Visuals: Perceptions of Visual Content Created by Professionals Versus Nonprofessionals
Limited studies directly address perceptions of photo source credibility, while several studies address issues adjacent to the concept. In 2018, Gayle, Wirzburger, Hu, and Rao used the Flanagin and Metzger (2000) scale to examine whether credibility perceptions of images were influenced by a label that stated the image was taken by a professional. They found that the professional label did not affect credibility ratings.
Researchers have noted clear visual differences between amateur photographers and professional photojournalists. These “accidental photojournalists” or citizen witnesses (Allan, 2014, p. 3; Andén-Papadopoulos & Pantti, 2013; Patrick & Allan, 2013). Unbound by the learned rules of what comprises a good photo, their aesthetic decisions vary. For example, amateur social media photographers tend to frame compositions less conventionally than professional photojournalists, and do not necessarily prioritize storytelling moments (Pantti, 2013). Amateurs often stand farther away from what they are photographing to avoid intruding than professionals do, and the subjects of their pictures are frequently centered instead of implementing the rule of thirds technically, the photos are much more often blurred (Barnhurst, 1994; Chalfen, 1987).
In a 2015 interview study by Allan and Peters, respondents were questioned about their perceptions of citizen photojournalism appearing in the news. Generally, they recognized the superior polish and technical savvy of professional photojournalism, but said that citizen photojournalists tended to portray lesser-known stories and alternative angles, lending to a sense of authenticity and adding a personal touch. An appreciation for the authenticity provided by imperfections has been echoed in others’ studies (Murray, 2008; Pantti, 2013.) That citizen photojournalists were on the scene when news happened was also perceived as an advantage. To some, the “truth value” of citizen photography was deemed greater than that of professionals because of its “rawness.” A few participants even perceived professionals as less credible, because it was perceived that they had a greater ability to stage or edit photographs.
Citizens’ perceptions of photojournalists manipulating photos was echoed by Mortensen (2014), especially with regard to staging, although the respondents in the study were limited to those considered citizen photojournalists. In a study of the same sample, Mortensen (2014) found that citizen photojournalists perceive photo blurriness and small imperfections to lend to a sense of authenticity. The respondents reported placing less value on traditional conventions such as good lighting, composition, and portraying the human element. Similarly, ordinary citizens who contributed visuals to the news–citizen visual contributors–were found to have greater credibility perceptions of citizen visual media than professional visual media in Chung and Jeong’s (2021) research.
Mortensen et al. (2021) compared viewer credibility perceptions of stock photographs and staff photographs, finding that stock photographs were perceived as less credible than those taken by a professional photojournalist. This was true on most statements, but respondents rated the stock photos more favorably on the statements that comprised what was termed “Journalism Professionalism,” or those that addressed the overall polishedness of the photographs: “a skilled photographer took this photo” and “anybody could take a photo of this quality” (p. 9). Noting the need for a measure of news photo credibility due to the inclusion of nonprofessional photographs in the news, those authors (Mortensen et al., 2021) adapted previous media-credibility scale statements and applied them to the context of photography that appears in the news. The resulting 18-statement, nine-dimension scale will be used in the present study. The authors defined photo credibility scale in journalistic contexts as “an individual’s assessment of the veracity of a photograph that appears in a news context” (p. 7). The dimensions are Authority, Accuracy, Coverage, Completeness, Currency, Journalism Professionalism, Objectivity, Trust, and Photojournalism Professionalism.
Summary and Research Questions
Scholars of journalism and media have long studied the factors influencing how credibility, trust, and believability are conveyed to and received by the audience. The popularity of social media has led to a body of research on what contextual, personal, presentational, and visual factors influence perceptions of credibility and trust.
Yet, it has not been fully quantified how the contemporary reality of visual interplay and cross-pollination between professional and amateur visual authorship, presented in news and non-news contexts, affects audience perceptions of credibility. In a media ecosystem awash in imagery and information—and sometimes lies and disinformation—a crucial goal of scholars and news outlets alike should be to understand how an audience judges an image to be credible in the real-world media ecosystem.
The following research questions were asked:
Research Question (RQ1). Is there a difference between perceived image credibility in images created by professional photojournalists versus those created by amateurs, even controlling for the context in which the image is presented?
Research Question (RQ2). Does the interaction of these authorship and contextual factors affect credibility perceptions in the audience?
Methods
Design of this online credibility study proceeded in three steps. The first was to choose stimulus materials. News events that were well-documented with both photojournalistic and Instagram imagery in 2020 and 2021 were considered. The goal was to find images taken by professionals and amateurs that visually represented the same news story in a similar manner. Visually, the amateur and professional pair would need to represent the same situation with the same primary actor(s) participating in the same primary action(s).
Three recent news events were chosen by a member of the research team with experience as a professional photojournalist: protests over the death of Breonna Taylor in a police shooting in Louisville, Kentucky, in 2020, the pro-Trump insurrection at the U.S. Capitol on January 6, 2021, and the Dixie wildfires in northern California over the summer and fall of 2021. The majority of amateur images published in newspapers depict “hard news” stories (Nilsson & Wadbring, 2015). The Capitol insurrection and the death of Breonna Taylor undoubtedly have a political valence, and arguably the climate change-inflected wildfires in the American West do as well. Yet, the advantages of choosing stories of high emotional intensity are that the audience member is more likely to care about the image, regardless of their political opinion. Thus, the viewer is more likely to have an emotional investment in the credibility (or lack thereof) of an image.
Seven images taken by staff photojournalists were chosen after extensive internet searches on wire service feeds and photojournalism contest-winning images. Once the professional images were chosen, the matching process began with a keyword search of publicly available Instagram photographs from nonprofessional users on that platform–as well as on Twitter, where cross-posted Instagram images were also searchable. The team narrowed seven potential pairs of stimuli down to the three pairs most balanced with similar situations and similar actions of the primary subjects of the photos (Appendices C to E). For the present paper, the researchers were not interested in the survey measuring differences in political perception of the news events depicted.
Qualitative aesthetics such as composition were a consideration in the search and the pairings, and a balance was sought. If the professional and amateur photographs appeared functionally identical, there would be little to study with RQ2. But if the images veered too far apart aesthetically, there would be no authoritative way to pinpoint what might have influenced a respondent’s credibility judgment.
Professional photojournalists pursue technically expert, well-lit and well-composed imagery as part of their storytelling mission (Kobre & Brill, 2017). Photojournalists cite professional compositional framing and correct exposure techniques as more definitional of photojournalism than amateur photographs do—although professionals do not express active aversion to the less technically expert and non-objective practices more consistent with amateur photography (Mortensen, 2014).
In the case of the study stimuli, the photojournalist’s attention to high-quality exposures and compositional techniques were apparent. In the professional Dixie fire photo, the photojournalist used a zoom lens to photograph a burning home at night, and successfully created a high-quality exposure out of a tricky lighting situation (Appendix D). At a protest over the shooting death of Breonna Taylor, the photojournalists used a wide-angle lens, a low angle, and the rule of thirds to bring compositional purpose to an aesthetically busy scene populated by multiple protesters (Appendix C). All chosen professional images were horizontal and presented in the more traditional 3:2 aspect ratio format.
Conversely, all three images taken by amateurs and posted on Instagram exhibited the aesthetic frumpiness common in amateur photography. While there are certainly talented amateur photographers on Instagram and technically exceptional photographs posted by nonprofessionals, researchers have noted quantifiable differences in the median visual approach of amateur photographers. For example, amateur social media photographers tend to frame compositions more unconventionally than professionals, and do not prioritize decisive storytelling moments (Pantti, 2013). Amateurs often stand farther away from what they are photographing to avoid intruding than professionals do, and the subjects of their pictures are more often centered in the frame and blurry (Barnhurst, 1994; Chalfen, 1987). To create a clearer visual contrast in this study, amateur images were chosen that embodied some or all of these findings about how amateurs take photographs.
Specifically, all three social media images were framed more unconventionally than their professional counterparts. Bright lighting in all three of the study’s chosen amateur images was overexposed, creating patches of blurry, blown out, or grainy visual effect. All three Instagram stimuli also featured the visual suggestion that the picture was taken by a participant in the news event. A photographer being self-expressive within the image they produce—often even appearing in the frame—is a systematic trope of amateur photography that Pantti (2013) calls “embodied collectivity.” In the January 6 Instagram photo, the officer’s gun is pointed at the protester/photographer (Appendix E); a viewer sees the Dixie wildfire raging through the doorframe of the photographer’s home (Appendix D); the Breonna Taylor protest photo is taken from amid the group of protesters and strongly implies that the photographer is a part of the demonstration (Appendix C). This is generally an approach different from professional photojournalists, who are taught to be witnesses and separate themselves from what they photograph, with NPPA (2017) Code of Ethics saying photojournalists should “not intentionally contribute to, alter, or seek to alter or influence events.”
Once the image stimuli were chosen, a generic online news presentation was created to feature the chosen images. Text that was as neutral, short, and descriptive as possible was written to be presentable as either the headline in the professional context or the Instagram caption in the social context. Research has shown that how a photograph is labeled can affect the cognitive and emotional responses of the viewer (Mendelson & Papacharissi, 2007), so care was taken to keep textual and design variables as consistent as possible within the presentation stimuli.
Usernames on Instagram were fabricated to be generic, and the avatar photographs were chosen from stock photos in which no individual facial features could clearly be identified for reasons such as appearing in silhouette. For the presentational mockups, username and number of likes remained identical in the social media contexts, regardless of whether the image was taken by a professional or an amateur. Identical blocks of Instagram comments from the original post (in social media contexts) and identical blocks of news copy from the original article (in professional news contexts) were added, but blurred illegibly. With the templates prepared, each of the six images (three news stories, with one professional photojournalistic picture and one amateur picture each) were placed in the professional context template and the modified Instagram feed template. Qualtrics was set up so that each pair of images would present one randomized condition per news story to the survey respondent, along with a series of questions pertaining to image credibility (Appendix F). Each survey respondent saw three images total to complete the survey, and no survey respondent saw both the professional and amateur image of the exact same news event.
Survey
This study made use of the Photo Credibility in Journalistic Contexts scale developed by Mortensen et al in 2021. This 18-statement, 9-dimension reliability scale was pared down to eight statements. Since participants had to take the survey three times, the authors abbreviated the scale in an effort to avoid survey fatigue. The researchers went through the original scale to discern one item from each dimension that was most relevant, and then worked to balance positively-worded and negatively-worded statements. Some statements were adjusted very slightly to fit the research at hand, though the meaning of the statements remained the same. No items were used from the dimension of the original scale termed “Photojournalism Professionalism,” as that metric was designed specifically for photographs that included people—and two of the photos chosen for this quasi experiment did not contain people. However, a separate “professionalism” metric remained in the current survey. The resulting scale did not yield reliability; thus, items are analyzed individually.
The statements ultimately used were (reverse-worded items marked with an asterisk):
This image shows an event occurring just like it did in real life.
*The information in this photo is not reliable.
Viewing this photograph helps me understand the (insurrection, Dixie Fires, Breonna Taylor)
This image tells the whole story about the topic
*This image could accompany many other posts
A skilled photographer took this photo
*The image tries to make me think a certain way about the topic Trust: I trust this photograph.
Data Collection
After IRB approval, the survey was designed and hosted on Qualtrics. The functionality of the survey was pre-tested for desktop, laptop, and mobile phone devices. Respondents were recruited from Amazon Mechanical Turk (www.mturk.com). The survey was conducted the week of March 4, 2022, and 4,486 individuals responded.
Results
A series of independent sample t-tests were conducted to determine differences between variables. To ensure that the stimuli themselves were not an unexpected source of variance, we ran an ANOVA to compare the means of the eight dimensions between the three different stimuli. The test showed a significant result in the Accuracy dimension, F (2, 2,987) = 6.0, p = .003, suggesting that the content of the stimuli generated unexpected variation. The Tukey post hoc test indicated the significance existed between the Insurrection images (M = 3.74, SD = 0.925) and the Dixie Fire images (M = 3.83, SD = 3.83) and between Insurrection and the Brianna Taylor images (M = 3.84, SD = 3.84). The variation in the Accuracy dimension was small and no other dimensions showed unexpected differences. An ANCOVA was also conducted to determine if there were any interaction effects between independent variables. No significant interaction effects were found in any of the scale dimensions. Given these results, the conditions are compared wholesale rather than image-by-image to allow for a more clear and direct answer to the research questions.
To determine whether there were significant differences between the way respondents viewed the credibility of images taken by professionals and those taken by citizens. There were significant differences for the Professionalism dimension. Across both the news and social media context, professional photographs (M = 3.80, SD = .94) were perceived by respondents as slightly but significantly more “professional” than the citizen-taken photographs (M = 3.70, SD = 1.01), t (4467) = −3.56, p < .001. The effect size, calculated with Cohen’s d, was very small, d = −.11.
Independent t-tests were conducted to determine whether there were significant differences between the way respondents viewed the credibility of images presented within the context of a news website compared to images presented on a social media site. Including both professional and citizen photography, there were no significant differences.
Comparing just the images presented in a news context, two variables showed significant differences between professional and amateur photos: Authority and Professionalism. For the Authority variable, professional images viewed within a news context (M = 3.36, SD = 1.19) were viewed as less authoritative than citizen-provided photos (M = 3.46, SD = 1.14), t (2,278) = 2.15, p = .05. The effect size, calculated with Cohen’s d, was very small, d = .09. There was also a significant difference for Professionalism on a news context (M = 3.81, SD = .91) compared with a social context (M = 3.70, SD = .98), t (2,278) = −.283, p < .001. The effect size, calculated with Cohen’s d, was very small, d = −.12.
Photos taken by amateurs and presented in a news context were contrasted with photos taken by professionals and presented in a social context. Across all dimensions, a significant relationship was found only in the Professionalism variable, t (2,259) = −2.47, p = .013. The effect size, calculated with Cohen’s d, was very small, d = −.10. Viewers rated the professional images in social context (M = 3.80, SD = 0.98) as more professional than the amateur image in a news context (M = 3.70, SD = .96).
Photos taken by amateurs presented in a social context were compared to professional images presented in a news context. Again, the only significant difference was found in the Professionalism variable, t (2,165) = −2.550, p = .011. The effect size, calculated with Cohen’s d, was very small, d = −.11. Amateur images in social context (M = 3.70, SD = 1.03) were viewed as less professional than professional images in a news context (M = 3.81, SD = .91).
Discussion
In a media ecosystem increasingly dominated by visual mediums—and increasingly rife with misinformation—how news consumers perceive the credibility of journalistic imagery is an important question. Using textual prompts, a study by Sterrett et al. (2019) found that the identity of a sharer was far more significant to credibility than the identity of the news source itself. But the credibility of visual journalism in these contemporary digital ecosystems has been a less explored topic.
Because there is fluidity in who authors an image depicting a news event—and in what context those images might appear—it is necessary to consider how these variables in combination affect the audience’s perception of the image’s credibility. In some studies, audiences that perceive a source and message to be more credible have been found to engage more with the news articles (Nedelcu & Blaban, 2021; Sterrett, 2019). However, other research, such as a 2021 study in Nature, found no correlation between a reader’s perception of the accuracy of a news article and their intention to share it. This was true even when participants expressed the belief that they should only share accurate articles (Pennycook et al., 2021). While this study’s survey results provided findings related to credibility perceptions, it did not attempt to connect those perceptions to sharing intentions or story engagement.
Beyond the news literacy of the polity, there are practical implications for understanding how image credibility differs depending on the source and presentation of a particular newsworthy image. Instagram has more than 2 billion regular users (Rodriguez, 2021), while even the most optimistic estimate of full-time professional photojournalists in the United States would number only in the tens of thousands (U.S. Bureau of Labor Statistics, 2021). With such a yawning imbalance between the number of amateur and professional images being produced, editors and publishers would be wise to consider how the audience perceives the credibility of amateur images and how to present these images in a credible context. Likewise, the question of how social media users judge the credibility of news content presented alongside a potpourri of other visual content on a site like Instagram is a question news outlets need to consider.
Scholars have recognized the singularity of sharing news content through an individual and/or platform that did not author the content originally. Sundar and Nass (2001) termed this the “selecting source” and Sterrett et al. (2019) used the word “sharer,” finding the identity of sharer has more effect on perceived trust and potentially engagement with a post than the identity of the original source. The results of this survey support this finding by showing that works of visual journalism from established mainstream sources do not have any special monopoly on credibility from the audience’s point of view.
Furthermore, in this survey respondents indicated that professionally authored images added almost no value to the credibility of the story being told. A professional image presented in a journalistic context rated essentially as credible as that same image presented in an Instagram feed. An image taken by an amateur photographer and posted to Instagram neither gained nor lost credibility when presented in a news context.
This was true even though respondents rated professional images higher in terms of authority, photographic professionalism, and journalistic professionalism as compared to an amateur image taken of the same event. The data suggest that professional photojournalism remains able to signal its own singularity amid the flood of imagery coming across cell phone, tablet, computer, and television screens. Crucially, however, the respondents to this survey did not indicate that this singularity added value to the image’s overall visual credibility.
Where does this finding leave visual professionals? Like all journalists, photojournalists are navigating a media ecosystem where “truth is an outcome of collective sense-making rather than unilaterally decided by newsrooms” (Waisbord, 2018). News audiences exhibit an open-mindedness in their encounters with news imagery, regardless of who authored an image or in what presentational context that image appears. Artificial intelligence only promises to add more news-adjacent imagery created outside the ethical framework of visual journalism in the years to come, with recent research finding that “generative AI systems have the potential to produce images at the quality of a human photojournalist and the images may be able to trigger similar emotional responses to those that were taken and selected by human photojournalists” (Paik et al., 2023).
Alas there is no single tactic that can bolster credibility perceptions in a media environment saturated by technological changes, bad actors, and short attention spans. Nonetheless, professionally authored imagery is valuable even if an audience is willing to embrace the credibility of news imagery authored by nonprofessionals. Visual storytelling remains effective at activating emotion, engaging readers, and prompting meaningful public discourse (Midberry & Dahmen, 2020), and professional photojournalists by definition are qualified and available to work with other journalists and editors to tell newsworthy stories with depth, empathy and intelligence. No other cohort has the professional mandate and consistent incentive to uphold the premise that images should accurately depict the accuracy of a visual moment in service of expressing a deeper truth.
On the scale of individual practitioners, these results suggest that photojournalists like Marcus Yam are wise to bring their work and their professional voice into the cacophony of social media spaces. Sharers themselves are part of the trust and credibility relationship a news consumer has to the news, with recent findings like “the sharer and the source do not interact to impact trust or engagement” (Sterrett et al., 2019). Framed more positively, Marcus Yam’s important visual work from Ukraine and elsewhere can potentially catalyze sharers across the world without impacting the credibility of the work or the news organization itself.
These results also suggest that news organizations might further broaden their definition of the photojournalism they publish to include more imagery created and published by amateurs. If the organization performs the work of finding and verifying amateur imagery that visually informs its audience, they are providing a useful journalistic service without damaging their own institutional credibility.
The inherent nature of quasi-experiments is one of the limits to this research, and there are also complexities in how images, text, and pre-existing biases might interact for any given specific representation of a real-world news event. An additional limit to this study, as scholars have noted, could be that focusing on still images in isolation does not fully mimic the visual ecosystem of online news websites or social media platforms. Those platforms and websites often include elements such as video, screengrabs, or visualizations (Keith et al., 2010).
Two of the three test pairs (Breonna Taylor protests in Louisville and the January 6 insurrection) were purposefully chosen as newsworthy events with high emotional salience. That’s because regardless of the politics of a particular respondent, those images were likely to prompt attention and consideration. The authors thought that respondents were likely to have an emotional response to these stories and thus would be more likely to feel invested in the credibility of the images. Prompting this emotional investment was worthwhile in order to gather relevant data, even though it likely contributed to the small but significant unexpected variation in the accuracy dimension between the three pairs of images, as detailed previously. Those three individual pairs of stimuli were assessed very slightly differently, potentially complicating these findings on the margin.
This experimental choice mimics the highly competitive online and social media ecosystem, where the fight for attention is paramount and often overshadows arguments about accuracy and truth (Phillips, 2018). But an emotional response might invite questions about other aspects of credibility perception that this study did not consider. This survey’s goal was not to map the connections between politics and imagery, but rather present survey respondents with the kinds of politicized news events so often present in both professional and social media contexts.
While this study offers insight into viewers’ perceptions of the credibility of visuals online based on source and medium credibility, it offers no insight into whether these perceptions will ultimately influence the sharing or spreading of these visual stories. As noted, research into this area is inconclusive on the exact relationship between credibility and sharing intentions (Nedelcu & Blaban, 2021; Pennycook et al., 2021; Sterrett et al., 2019).
Fruitful future research could work to further explain the aesthetic, informational, and presentational factors that influence how an audience rates the credibility of news imagery. In so doing, this research could help news organizations more productively embrace the reach of social media platforms and images created by amateurs in the ultimate service of better informing the public visually.
Footnotes
Appendix
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from the College of Information and Communications at the University of South Carolina.
