True or False? Viewer Perceptions of Emotional Staff and Stock Photos in the News

The phenomenon of untruthful photography appearing in journalistic contexts flies in the face of the photojournalistic ethical imperative to truthfully depict and document the emotion inherent in the human experience. Heavily manipulated images, a concern that rose at the turn of the century, are joined with AI-generated images, deepfakes, and stock image banks, creating a slew of potentially deceitful imagery in the media. Creative stock photos, the subject of this study, are not truthful and do not depict genuine emotion. While editorial stock image banks do exist, the subject of the present study are those with staged scenes, created with the intent of multiple uses, and customizable to various needs. These stock photos are often used as a means of acquiring low-cost visuals (Adobe, Citation2023). Found historically primarily within the persuasive media realm, these stock photos also sometimes appear in the news (Frosh, Citation2001) and social media sites. Photographs taken by staffed photojournalists, on the other hand, are shot with a professed commitment to ethical values (Elliott, Citation2003; Fahmy et al., Citation2014). What separates a journalist from other content creators, Singer (Citation2007) say, are “ethics or norms, as well as the principles that underlie those norms” (p. 47).

A most esteemed journalistic ethical principle is truth (Kovach & Rosenstiel, Citation2021). Untruthful visual news can be harmful to viewers who rely on it to make informed decisions; it also can ultimately damage the credibility of the news industry (Elliott, Citation2011; Silva & Eldridge II, Citation2020; Self, Citation2014). Credibility encompasses other variables related to truth, including accuracy, completeness, objectivity, and trust.

Depicting the emotion involved in the human experience is a second hallmark norm of photojournalism (Lester, Citation2018; Newton, Citation2013; Rössler et al., Citation2011). Emotion is characterized by valence (negative or positive tone) and arousal (high intensity or low intensity; Bradley & Lang, Citation1994). Photos with a negative valence are the most common and most rewarded in photojournalism (Godulla et al., Citation2021; Singletary & Lamb, Citation1984; Shoemaker & Reese, Citation2013). Negative photos can increase outcomes such as attention, memory, assessments, and news judgment (Calvo & Lang, Citation2004; Emmett, Citation2011; Knobloch et al., Citation2003; Martin-Kratzer, Citation2005; Rössler et al., Citation2011; Smith-Rodden & Ash, Citation2012). For example, a March 2022 photo of rescue workers evacuating a pregnant woman from a Ukrainian maternity ward after a Russian attack impacted how Americans visualize and perceive the war (Bracaglia, Citation2022).

Given these principles, using stock photos in a news context has caused concern among ethicists and journalists alike. When the Poynter Institute suggested the use of free stock photos by journalists to use in articles in 2018, members of the National Press Photographer’s Association, Associated Press Photo Managers, and The Kalish Visual Editing Workshop expressed “profound disappointment” (Poynter Staff, para 1). They cited the potential for accidentally using unlicensed images and the risk of obtaining an image that has been manipulated. They further pointed out that the interaction between an image and a story may lead to false conclusions. Finally, they argued that audiences recognize and appreciate the quality of professional photos. Much of the debate about using stock photos revolves around the question of whether audiences can detect stock images as such. Sonderman (Citation2013) notes, for example, that unless stock photos are blatantly obvious, “at best” the reader is uncertain about whether a photo is stock (para 6; see also Staff, Citation2018). Beyond occasional anecdotes (e.g. O’Rourke, Citation2022; Putterman, Citation2021), there is not currently a solid understanding of whether and in which what ways audiences are being deceived by stock photos, a potential harm to viewers and photographed alike (Elliott, Citation2011). In March 2022, users incorrectly scrutinized the authenticity of a photo of Joe Biden on Twitter (now X) due to the green foliage in the background (O’Rourke, Citation2022). But viewers were correct in their detection of the falsehood of an image published by the New York Times when one user posted “SHAMEFUL – The New York Times publish a 2015 stock photo of a young girl claiming Israeli forces killed her during last week’s war with Hamas.” In this case, the photo was from 2015 and posted incorrectly with a similar story in 2020 (Putterman, Citation2021), leaving the impression that, as one poster noted, the girl to seem to be “killed twice.”

Truth and emotion are distinctly photojournalistic axioms, but blatantly false photos can and do exist in news contexts. While scholars and news professionals recognize the ethical concerns, little is known about audience detection of and deception by false images appearing in a news context. Learning whether and in what ways audiences can perceive the difference between an untrue stock photo and a truthful staff-created photo has implications for the credibility of professional photojournalism and ramifications for an industry with both an historical and contemporary devaluation of photojournalists (Klein-Avraham & Reich, Citation2016; Zelizer, Citation1995). The implications of this study can be extended into areas such as ChatGPT and other artificial-intelligence created image, as well. Learning about the way audiences can or cannot detect subtle cues indicating visual falsehoods may help justify the use of professionally-shot news staff photographs. Further, learning about the ways audiences are being deceived or not will help inform news organizations’ future ethical decision-making around the use of stock photos.

The purpose of this study is to examine the effects of emotional valence and source upon viewer perceptions of a stock and staff photo in a news context. A two (source; stock versus photojournalist) by two (negative versus positive valence) quasi-experiment is undertaken. Replicating and augmenting previous studies, viewer perceptions are measured with five dependent variables: 1) credibility perceptions using the news-photo credibility scale; 2) self-reported arousal level, 3) emotional valence perceptions, 4) fixation duration, and 5) fixation count.

Evolving state of news visuals

The photojournalist layoffs of the last couple decades caused by evolving business models and a long history of photojournalists viewed as second-class newsroom citizens has left visual vacancies in the news (Brennen, Citation1998b; Gade, Citation2011; Klein-Avraham & Reich, Citation2016; Lydersen, Citation2013; Zelizer, Citation1995). As online sources of visuals have become hugely abundant and digital technologies afford nearly everyone to take a technically-sound photograph, photojournalists are viewed as a relatively disposable newsroom asset (Brennen & Brennen, Citation2015; Klein-Avraham & Reich, Citation2016; Lowrey, Citation2002). Some have said that it seems almost any type of visual will suffice, including citizen-shot photos, file photos, or generic stock photos (Klein-Avraham & Reich, Citation2016; Sonderman, Citation2013). Sometimes news sites “just need a photo, any photo, to color an otherwise gray slate of text” (Sonderman, Citation2013).

Stock photography

Photographers of non-editorial stock images create photos that anticipate the future needs of various media outlets with the ultimate goal toward breadth of applications. Getty Images and others offer both stock and journalistic imagery, which it refers to as either “creative” or “editorial” (Gaston, Citationn.d.). The “creative” style of stock photos is the subject of the present research. Since its 1974 inception, the stock photo industry has grown greatly (Frosh, Citation2001). In 1995, Getty photos created the first online photo bank, increasing availability (Machin, Citation2004). The Global Stock Photos Market Report noted $2 billion in revenue in 2020 with a projected 7% annual growth through 2026 (Expert Market Research, 2021). The nature of the industry has led to the stereotypical generic stock photo: a “glossy, formulaic, multi-purpose representation” (Frosh, Citation2001, p. 30). “Good” stock photos are said to have vivid colors, perfect white balance, a sharp subject, typically with a blurred background, low ISO for crisp focus, and a horizontal orientation (Wallon-Hárs, Citation2023). As for the content, stock photos have been described as having “a weird, posey cheesiness often unified by hilariously bad photo editing” (Baliūnaitė, Citation2022). As Baliūnaitė (Citation2022) claims, “When you see a stock photo, you know it’s a stock photo” (para 1, italics in original). And yet, stock photos have been called “the visual backbone” not only of advertising, branding, and publishing, but also “journalism” (Aiello, Citation2016, n.p; Frosh, Citation2001). In more recent years, stock photographers have moved toward a more authentic-appearing aesthetic (Aiello, Citation2016). The rise of smartphones has driven a trend toward less technically perfect and sterile shots (“7 photographers,” 2018). These conceptual and visual changes can be ethically worrisome to those in the journalism industry, with lines blurring between posed and authentic.

Photojournalism ethics and norms

While content is abundantly available online, what distinguishes journalism from other activities related to gathering, editing, and disseminating information to the public is its ethos (Singer, Citation2007). Photojournalists working in a newsroom traditionally profess to abide by ethical standards and are guided by deep-seated principles (Fahmy et al., Citation2014; Silva & Eldridge II, Citation2020). News photos have been credited with having the “extraordinary power to move readers’ consciences and prompt action (Santos & Eldridge II, Citation2020, p. 2), though some scholars challenge the assumed impact of photos (e.g., Perlmutter, Citation1998). Elliott (Citation2003) advise thatbecause visuals have the emotional power to educate, entertain, and persuade, there is a responsibility put on each photo created for the public (p. 24).

Emotion

Photojournalists-in-training have been coached to capture the emotion involved in the human experience since virtually the development of the profession (Costa, Citation1950). Emotion is defined in psychology by the dimensions of valence and arousal (Barrett et al., Citation2007; Lang & Bradley, Citation2010). Textbooks consistently urge students to both capture emotion and thus to evoke emotion among viewers (e.g. Arthur, Citation2016; Horton, Citation1989; Kobré, Citation2011; Skinner, Citation2010). Intimate depictions of emotion is a key consideration used to judge contests, with more emotional photos winning more awards (Lough, Citation2021; Singletary & Lamb, Citation1984). Professional photographs, overall, show more emotion than nonprofessional photography (Greenwood & Thomas, Citation2015; Mortensen, et al., Citation2023). Editorial departments, too, tend to choose photos that show greater emotion (Rössler et al., Citation2011). Emotional photography is necessarily intertwined with notions of truth; the dramatization of emotion in stock photos are explicit falsehoods, infringing on photojournalism ethos.

Research questions, and hypotheses

The literature shows that truth and emotion are photojournalistic axioms with ethical implications, but very little is known about perceptions of photos that are traditionally neither truthful nor emotional: multi-purpose stock photos. Related studies have suggested that photo source can affect audience perceptions, that audiences often react more strongly to negatively-toned photos, and that emotional photos are sometimes associated with emotional reactions. Learning whether viewers perceive the difference between a stock and staff photo will begin to help shed light on the degree to which viewers might be deceived, ultimately harming industry credibility. In this study, we respond to a call for increased empirical research exploring perceptions of news photographs (Müller et al., Citation2012; Smith-Rodden & Ash, Citation2012). Specifically, we look at participants’ perceptions in a 2 (source; stock versus photojournalist) x 2 (negative versus positive valence) quasi-experiment, measuring: 1) self-reported arousal level, 2) self-reported valence perceptions 3) credibility perceptions, 4) fixation duration, and 5) fixation count.

Replicating and augmenting Mortensen and colleagues (2021), we hypothesize:

H1:

Respondents will perceive staff photos as more credible than stock photos a) overall, b) with positive photos and c) with negative photos.

Adding to self-reports of credibility perception and moving toward psychological, quantitative measures we ask:

RQ1:

Is the source of the photo associated with different a) self-reported arousal-levels, b) valence perceptions c) fixation duration, and d) fixation counts?

And adding to literature most often found within the psychology realm and primarily assumed by the news industry, we ask:

RQ2:

Is the valence of photos associated with greater a) credibility perceptions, b) self-reported arousal levels, c) valence perceptions, d) fixation duration, and e) fixation counts?

Method

Quasi-experimental design and procedure

A two (valence: positive vs. negative) x two (photo source: stock vs. professional photojournalist) mixed-factor design quasi-experiment was conducted. The quasi-experiment was conducted in a laboratory setting equipped for eye-tracking research. What makes the design a quasi-experiment is the nature of photos themselves. We did not isolate a single variable, such as a cutline. Rather, we were interested in whether participants could perceive the nuanced differences between the two types of photos. Other variables were necessary and unavoidable. Participants were invited and informed about the study via SONA, an internal recruitment system at the research institution. Participants were sent a confirmation e-mail that included an informed consent letter. Participants were asked to sit in a separate room behind a 22-inch monitor at a distance of 60 cm. A Tobii X3–120 eye-tracker was attached. The researcher sat at a separate station outside of the room. Eye-tracking data were collected using a sample gaze rate of 120 Hz per second, and gaze samples were recorded for eye fixations of about every 8 milliseconds. Once calibration was completed, participants were instructed to look at either a positive or a negative stimuli in random order for 30 seconds each that appeared in the form of an editorial regarding opioid addiction (See Figures 1–4). Participants were randomly assigned to view the two different valenced photos taken by either stock photographers or professional photojournalists. They did not know which photo source condition they were given, as determining whether the photo itself affected perceptions was a purpose of the study. Participants were then asked to complete a questionnaire about their emotional responses toward each photo and the perceived photo credibility, as well as demographic information. Participants were thanked and compensated with a $10 gift card.

Figure 1. Staff photo, negative tone.

Display full size

Figure 2. Stock photo, negative tone.

Display full size

Figure 3. Staff photo, positive tone.

Display full size

Figure 4. Stock photo, positive tone.

Display full size

Results

Hypotheses 1 proposed that respondents would perceive staff images as more credible than stock images a) overall, b) with positive images and c) with negative images. All three parts of the hypothesis were supported. The two-way Univariate Analysis of Variance (ANOVA) showed that a) staff photos (M = 3.20, SD = .56) received significantly greater credibility perceptions than stock photos (M = 2.92, SD = .63); (F(1,160) = [11.09], p < .001, p2 = .07). This was true for negative and positive photos, with respective mean differences of .27 and .29. Staff photos are perceived as more credible than stock photos on almost all scale items, with two exceptions. To the Professionalism item, which includes the statements, “A skilled photographer took this photo” and “Anybody could take a photo of this quality” (stated inversely), those who viewed stock photos were significantly more in agreement with the professionalism concept, with significant mean differences of .28 (negative) and 1.07 (positive) photos. Staff images are also perceived as being significantly less “current” than stock images on both negative (mean difference = .18) and positive valence conditions (mean difference = .16).

RQ1 asked whether the source of the image was associated with different a) different self-reported arousal-levels, b) different valence perceptions c) gaze fixation and d) gaze duration, and whether there were any interaction effects for each variable. The answer is mixed. Results showed no significant main effects, but some significant interaction effects.

Research question two asked whether the valence of images is associated with greater a) credibility, b) self-reported arousal levels, c) valence perception, d) fixation counts, and e) fixation duration. Results of the two-way ANOVA tests show that:

Discussion

This study sought to examine perceptions of staff and stock photos with different valences. The results of the study both confirm and contradict the literature. Overall, participants viewed staff photos as generally more credible, and this applied to both positive and negative valences. In both stock photos and staff photos, negative photos were perceived as slightly less credible. Stock and staff photos were similarly arousing, received similar valence perceptions, and were associated with similar gaze fixations and durations, overall. Interaction effects in terms of source were apparent, though, with stock photos being associated with greater effects in most instances. In terms of valence differences, negative photos were much more arousing and received greater fixation counts. Positive images, opposed to negative images, received much more positive emotional responses.

That participants’ perceptions of the credibility of staff photos were greater than stock photos was a finding consistent with Mortensen and colleagues (Citation2023) who also used the same scale to test credibility perceptions of stock and staff photos. This finding bolsters the argument made by members of professional photography associations NPPA, APPM, and The Kalish that audiences can detect between professional and other kinds of photos, spending more time with professional ones. As they say, “It’s not good enough to simply have a photo with your story. The content of images and captions matter” (Staff, Citation2018, para. 9). The mean difference of .28% points was modest but significant. Overall, participants perceived the genuine emotion portrayed truthfully in photojournalism as opposed to the faux emotion acted out in stock photos. The pattern of credibility perceptions was the same with both valence conditions. This replication adds validity to the scale and reinforces the notion that audiences perceive greater overall credibility of staff photos. Knowing that audiences can detect some differences between staff and stock photos supports the argument that quality photojournalism matters. The one scale item in which the stock photo condition was rated as more credible in the Mortensen and colleagues (Citation2023) study – professionalism–was also rated more credible in the stock condition in the present study. Viewers’ perceptions that stock photos are more “professional” in this regard might reflect the squeaky-clean, technically sleek nature of traditional stock photos (Frosh, Citation2001; Wallon-Hárs, Citation2023). Stock photos are in fact taken by skilled photographers who have the time to stage the photo and get the lighting and composition perfect in a way that on-staff photojournalists might not, thus lending to what viewers may perceive as a “professional” appearance.

Importantly, the one statement with no significant difference concerned whether the photos were relevant and specific to the story. While the fact that only one statement had no significance may at face value seem to imply that there is very little problem with using stock photos, this particular statement is telling. The insignificance suggests that while audiences seem to note differences between the two types of photos, they may not understand that stock photos were not taken specifically for the story at hand. This contradicts what is known about stock photos – they are created to accompany many contexts (Aiello, Citation2016). This suggests that viewers are being deceived in perceiving some stock photos in the news as taken specifically for the news story. The Poynter Institute (Staff, Citation2018) notes that not fully considering the interaction between an image and the story can lead to false impressions about a subject, a point that should help guide newsrooms’ policies on the use of stock photos and inform education in photojournalism. In the same way the profession and higher education warn against staging and altering photos and when minor alterations might be deemed acceptable, there should be serious discussion as to when, if, and in what manner stock photos should ever be used in a news context.

The findings regarding valence are somewhat consistent with literature and with the journalism industry’s previous values and patterns, but more mixed. The journalism industry’s pattern toward the negative was suggested to be shared by audiences in the present study (e.g., Godulla et al., Citation2021; Lough, Citation2021; Singletary & Lamb, Citation1984). Negative photos generated much higher arousal, with a large mean difference of .77. This was the strongest difference in the study. Studies have consistently shown that negative photos are rewarded in the news industry (e.g., Lough, Citation2021), and this study shows there is a payoff for doing so. However, negative photos did not drive greater attention as the literature would suggest (Knobloch et al., Citation2003; Rössler et al., Citation2011), except in the case of the stock photos, wherein they had greater fixation counts and duration. As these stock photos were dramatized, it seems they may be more attention-grabbing than the staff photos. The ability of stock photos to gain attention does not compensate for their ethical breaches. Interestingly, positive photos – regardless of source – were perceived as more credible than negative photos. One explanation is also a limitation of the study: given the quasi-experimental photographic study design, the stimulus materials were not identical with one isolated different variable. The positive photos in this study were in color, while the negative photos were black and white, for example. Another explanation is that the negative photos were viewed as sensational, thus less credible. While audiences might be drawn to the negative photos, the findings suggest that can detect at least some of their untruthful nature. In the case of the negative photos, both portray a woman injecting heroin, but the stock photo shows slightly more face, the needle is slightly more prominent, and it should be addressed that the clothing of the woman on the stock condition is more revealing. While it might be interpreted that the differences in the stimulus materials caused the results, it should be noted that the well-thought-out composition is afforded through the staged photos in the stock condition, but not the staff condition.

These results are difficult to interpret, but the researchers must concede that some of the perhaps unexpected results are likely due to the stimulus materials. Every step was taken to find comparable photos, but the nature of this kind of visual communication research is that it is impossible to isolate a single variable while using real-world photos. If the photos were identical except for a cutline, it would be impossible to gauge whether viewers actually perceive the different sources. Given this complication, the findings of this study are not generalizable but rather one piece of evidence in the pursuit of audience perceptions of stock or otherwise untruthful imagery. We urge more studies to replicate the present one and help control for the varying nature of photos, however slight. Another limitation of the study worth mentioning is the sample; only undergraduates at a single university participated in the study. This was a necessary limitation given the in-person nature of the lab setting.

The insignificant results are telling, as well. Stock and staff photos received similar valence perceptions, overall, which is one measure of emotion. In other words, they did not foster stronger negative or positive perceptions, overall. The two sources also fostered similar main effects for fixation duration and fixation, suggesting they draw similar levels of attention, though there are nuanced interaction effects with valence, as discussed above. These similarities, again, show a level of deception among viewers who see stock photos in the news.

This study begins to reveal some potential ethical implications of generic stock photos in the news. The findings are mostly positive for professional photojournalists and their viewers in that they hint at the importance of photojournalism. At the broadest level, they suggest viewers recognize the genuine, emotional nature of professional photojournalism when compared with stock photos. Still, more nuanced findings show that in some ways, viewers are being deceived and perceiving false images as truthful. Specifically, audiences do not necessarily understand that the photograph is not taken specifically for the story at hand, even though they can detect other visual differences between stock and staff photos. This deception has the potential to harm viewers trying to make sense of a story through the combination of words and visuals, the shaky credibility of the news industry, and – as is was the case with the New York Times anecdote – the subjects of photos. The initial results should give pause to an industry quick to dispose of photojournalists first when budgets are tight. The appropriateness of using stock photos in the news merits serious contemplation amongst professionals and educators. Future research into false imagery in the news could extend the present research into the arenas of artificial intelligence and deepfakes.