Visualizing the solution: An analysis of the images that accompany solutions-oriented news stories

Solutions journalism

Solutions journalism is a style of reporting in which journalists critically examine possible solutions to social problems. The trend emerged in a 1998 Columbia Journalism Review article that described a rise in reporting on ‘what works’ by known news organizations such as the Los Angeles Times, The New York Times Magazine, and ABC News (Benesch, 1998: 36). In recent years, solutions journalism practice has surged, largely due to the 2013 launch of the SJN, an independent, nonprofit organization. The SJN (2017) defines solutions journalism as ‘rigorous reporting about how people are responding to problems’ (n.p.). By late 2017, the SJN had partnered with more than 90 US news outlets and hosted its inaugural summit bringing together practitioners, academics and others who share its mission.

Despite its momentum in the industry, scholars have only recently begun to conceptualize solutions journalism. The term solutions journalism first appeared in a journal article when Davies (1999) suggested it as a way to halt the corporatization of journalists in Australia, but the article did not conceptualize the practice. Solutions journalism was first fully defined in the academic literature more than a decade later within the broader concept of constructive journalism (McIntyre, 2015). Constructive journalism has been defined as ‘an emerging form of journalism that involves applying positive psychology techniques to news processes and production in an effort to create productive and engaging coverage, while holding true to journalism’s core functions’ and solutions journalism has been identified as one technique for practicing the broader concept of constructive journalism (McIntyre, 2015). Solutions journalism may have roots in older, similar forms of news such as civic journalism, but the specific term is distinct (McIntyre, 2015). Reporters in the civic journalism movement that peaked in the 1990s focused on motivating audiences to contribute to the democratic process using methods such as hosting town hall debates (Rosen, 1999), whereas the solutions journalists of today do not emphasize a goal of getting the public to participate in civic issues. Rather, they focus on shifting the framing of news stories from highlighting a conflict narrative to highlighting a solution narrative, focusing on how people are responding to a broader array of societal conflicts (SJN, 2017).

Scholars have also only recently begun to study solutions journalism’s effects. Curry and Hammonds (2014) conducted a quasi-experiment in which they found individuals who read solutions-based stories reported feeling more interested, optimistic and informed and more likely to share those stories than those who read conflict-based stories. Despite its encouraging results, participants in quasi-experiments are not randomly assigned, and thus internal validity is threatened. In a true experiment, with random assignment of participants, results indicated less support of solutions journalism (McIntyre, 2017), but did show solutions-based stories caused readers to report feeling more positive affect and more favorable attitudes toward the news article and toward a solution to the problem than did conflict-based stories. In another true experiment regarding solutions stories, researchers found consumers of a solution-based story reported more positive affect and greater behavioral intentions to read similar stories (McIntyre and Sobel, 2017). And most recently, in an experiment and two field tests, Curry et al. (2016) found individuals who read solutions stories reported higher levels of self-efficacy and optimism and spent more time on the webpage than those who read non-solutions stories, but they also had higher bounce and exit rates.

In this early stage of research on solutions journalism, scholars have attempted to define the concept and tested its effects in and outside of the lab. What has not yet been considered, however, are the visuals tied to solutions journalism stories. This study seeks to examine just that.

Impact of images

Photos add meaning and understanding to news stories and can influence the reader. This has been measured via framing and agenda-setting in political perception (Coleman and Banning, 2006), terrorism/fear (Fahmy et al., 2006), personal health risk (Gibson and Zillmann, 2000) and memorability of news in general (Graber, 1990). In each instance, the presence of visuals contributes to, and in some cases overwhelms, the message in the story. Indeed, we know information contained in the visual component of the story may be more memorable to the reader even if it conflicts with information presented in words (Coleman, 2010), though some evidence points to a more nuanced multimodal approach (Geise and Baden, 2015; Jewitt, 2009; Kress and Van Leeuwen, 2001; Machin, 2016). However, Powell et al. (2015) found when the visual and text message are presented together, regardless of message conflict, the image message drove behavioral intentions while the text message drove opinion influence. Behavioral intentions are a key finding of solutions journalism stories, lending credibility to the need to examine visual content included with solutions-oriented reporting.

While visual components of a news story convey particular messages to the viewer, it is important to place them in the appropriate multimodal context – they do not exist alone and are not the sole source of interpretation, though they do carry unique meaning potential. Indeed, as Barthes (1977) argues, even the image itself needs a linguistic anchoring in order to fully form an understanding. While a story written with a strong emphasis on solutions journalism can generate particular thoughts and actions from readers, questions of word/image congruency suggest perhaps an accompanying visual contradicting the solutions frame might modify these effects or at least create a different gestalt interpretation than from the story alone. The stages of processing multimodal information are complex and recursive (Geise and Baden, 2015), which can result in interpretations beyond the intended meaning. Before we can test these effects, we need a fuller understanding of what types of photos typically accompany solutions news stories. Current research on solutions journalism has left out the visual aspect, which is what we hope to begin to correct with this research.

In broad, direct effects, the inclusion of photos alone increases selection and interest of news stories, and ‘threatening images’ versus ‘innocuous images’ have been found to increase selection and interest even more (Knobloch et al., 2003). A threatening image can imply danger or directly depict harm or victimization – not what one may expect to find in a solutions journalism piece. Rather, the innocuous image, one showing restoration versus victimization, would likely be found, though that might reduce reader interest compared to a photo of whatever conflict the story’s solution addressed.

In addition, research on the latent meaning of photos has shown how subtle compositional cues in the image can affect the reader. Kobré (1999) describes how decorated photographer and editor Joe Elbert puts photos in one of four hierarchical categories: informational, graphically appealing, emotional and intimate. At the lowest level, an informational photo gives basic data to the viewer, the who/what/when/where/why. Next, graphically appealing photos are still basic in the information provided but are presented in a more visually interesting way due to compositional decisions by the photojournalist. Third, emotional images begin to move past visual interest and into photos that ‘cause the reader to feel something about the subject, not just intellectualize about the story’ (Kobré, 1999: 18). The final, and what Elbert calls ultimate goal, is intimacy. An intimate photo takes emotion to the highest level, where the reader truly feels a deep connection to the subject.

Further detail into image effects reveals how more subtle cues can change viewer perception. Berger (1981), Coleman (2010) and others have explored how non-verbal compositional and subject elements such as camera angle, camera distance, presence of people, eye contact and activity all can have positive, negative or neutral connotations. A close-up of a face signifies intimacy, whereas a wide shot showing the full body of an individual signifies more social distance between the viewer and the photograph. A photo from above the subject, looking down on them, implies weakness, whereas taking a photo from below the subject, looking up, signals authority (Berger, 1981). Showing people, and more importantly people in action versus non-action, are considered more positive (Coleman, 2010). Another way of viewing a photo includes physical touch between subjects. Goffman (1967) and Montagu (1971) reveal non-verbal messages are communicated between persons based on touch. This communication is extended to a third party, the viewer, as found by Henley (1973) and Summerhayes and Suchner (1978).

These small, latent meanings are important when the original topic of the photo is considered. A photo could have one intention topically but then be juxtaposed in latent meaning, as seen in Huxford’s (2001) examinations of how visuals can redirect readers from the intended message. A photo depicting a rape survivor as being strong can be nullified if the latent meanings depicting the survivor are negative.

Photos convey meaning to the reader, in subject, composition and placement, both directly and indirectly. The message transmitted to the reader in a photo can overwhelm the message in the text, and so it is important to include visuals in the study of solutions journalism. To begin understanding this line of research, we first create a foundation of what visuals exist in current solutions journalism practice by asking the following:

Data collection

A total of 1241 news stories and accompanying photographs obtained from the Solutions Journalism Network’s ‘Solutions Story Tracker’ were examined using content analysis.

The ‘Solutions Story Tracker’ contains real-world examples of solutions journalism. As of late 2016, this database included more than 1500 staff-curated items from a range of publications including legacy media such as The New York Times and websites such as Medium. This presents a unique purposive sample in that only stories already identified as using solutions journalism are included. For this study, the database fits perfectly into the goal of understanding the types of images included in solutions journalism stories. When non-text/non-journalism items were removed, a final N of 1241 stories remained for analysis. Due to the nature of the population, sampling was not deemed necessary and the entire population was included in coding.

Staff at the Solutions Journalism Network obtain story links for the Story Tracker three ways: those created through partnership with the Solutions Journalism Network, those submitted by outside members of the public, and finally, those found by staff at the Solutions Journalism Network (E. Tompkins, 15 September 2016, personal communication). Stories selected for analysis were viewed and coded first in the Story Tracker and then on the respective news site. Within each story, the text and the first photo (if present) was included in analysis. Wanta (1988) found the highest salience in the dominant photo, which we use to justify our decision to look solely at the first photo.

Variables

Success factors are tenets of effective solutions journalism and are used to operationalize how we identify a ‘solution’ in this study. More specifically, they are distinct methods critical to a response’s success or failure. Staff at the SJN developed eight top-level success factors and used them to tag stories in the story tracker (Table 1). The success factors were generated from analyzing hundreds of news stories to ‘see what tactics were critical to making a response work’ (T. Nelson, 28 November 2016, personal communication). These categories are not mutually exclusive and in most cases two or more are tagged with each story. Since there is currently no standard method of defining a ‘solutions photo’, the researchers operationalized the definition by asking if the image content reflected any of the success factors used in the story. For example, a news story about helping ex-cons re-enter the workforce is tagged with a success factor of ‘empowering people’. If the accompanying photo depicted this, such as showing former prisoners in a training program, then it would be coded as reflecting a success factor. Photos that don’t reflect a success factor could be neutral, such as a photo of a prison, or even problem-oriented such as a photo of an arrest.

OPEN IN VIEWER

Table 1.

Top-level success factors.

Success factor	Description
Attacking root causes	Solving problems by addressing the underlying factors. By attacking the root causes of a problem, changemakers treat the disease rather than just the symptoms. These solutions offer help to those in need, employ a comprehensive approach to problem-solving, and address the determinants that keep individuals and communities from succeeding.
Cultivating collaborations	Reaching across the aisle and rounding up the community to solve problems. These solutions depend on coordinating many actors, on a macro or local level, to achieve success.
Embracing the power of relationships	Building and leveraging human networks to solve problems. By building trust and personal relationships in a community, social entrepreneurs create change on a behavioral level. They also make use of positive peer pressure and social norming – showing people that certain peers are behaving well.
Empowering people	Building the confidence and problem-solving ability of communities. Working closely with the demographics they serve, changemakers identify communities’ assets and build programs controlled by locals. In so doing, they encourage resilience and turn ‘victims’ into problem-solvers.
Expanding access	Bringing products and services to new people, in new markets, in new ways. These solutions focus on providing services to those in need. By expanding marketing and distribution, reducing costs, and rewriting rules in pursuit of equality, changemakers broaden the reach of their innovations, extending access to all.
Overcoming the big challenges	Showing how to solve problems innovators face. Challenges are endemic to social entrepreneurship. These stories illustrate how changemakers are addressing those challenges, and whether they’re finding success.
Practicing human-centered design	Customizing and tailoring solutions to meet a person or a community’s needs. In developing solutions, changemakers employ human-centered design, meeting people where they are at, employing behavioral economics, and identifying cases of positive deviance.
Using creative financing	Harnessing the power of economics and investment. These solutions take advantage of money, not only in its ability to fund programs but also as a means of giving people autonomy and the power to solve0 the challenges they face in the ways they see fit. These solutions take advantage of the incentives, power, and leverage of finance.

Descriptions summarized from https://storytracker.solutionsjournalism.org/success_factors/attacking-root-causes.

This way, conflict between the image message and story message can be identified, such as a story about how a country has overcome the big challenges of war but has a conflict-based photo of soldiers fighting.

Photo captions and photo credits were also coded, if included. Similar to the image analysis, information in the caption was examined to see if it reflected any of the success factors used in the story. Coders also examined the credit line to determine who took the photo: staff (internal), wire service (external), member of the public (such as a freelance photographer) or other (such as provided by the organization the story is about).

Stories are also tagged with numerous topical issue areas. Seventeen top-level issue areas were collapsed to create seven ‘primary topics’ based on shared commonalities. Primary topics coded were health, education, community and economic development, public safety and affairs, environment, culture and other.

Photos were classified by type based on categories used in respected professional photo organizations: general news, spot news, feature, portrait, sports and other/unknown/unsure (National Press Photographers Association, 2016; Pictures of the Year International, 2016). They also were coded for being dominant/non-dominant on the web page, that is if they were ‘above the fold’, appearing on-screen when the link was clicked and before the text of the story.

Several variables were included to assess the general emotion of the photo, individual emotional indicators and physical touch. First, the image was examined as a whole and coded into one of the four levels of Elbert’s Emotional Hierarchy: informational, graphically appealing, emotional or intimate (Kobré, 1999). If the image only provided basic information without visual appeal (interesting camera angle, use of composition/framing, etc), such as a photo of a ribbon-cutting for a drug treatment center, then it was coded at the informational level. If the same subject matter was instead presented with clear attempt by the photographer to add visual appeal, such as shooting the ribbon-cutting from a low, dramatic angle to also include the building’s sign and the sky, then it was coded at the graphically appealing level. If emotional depth was added, such as instead showing a moving portrait of someone who had successfully completed rehabilitation, then the photo was coded at the emotional level. If the emotion in the photo was so strong that it caused a deeper connection to the viewer, such as showing an inside look at life on the streets of drug users, then it was coded at the intimate level.

Next, a ‘nonverbal expressions’ index (Coleman, 2010) was created using four variables identified in previous studies on images and emotions. Negative indicators were coded when the camera angle was from above, the camera distance was far away, the people (if included) were passive and eye contact was indirect. Neutral indicators were coded when the camera angle was straight on and the camera distance was medium. Finally, positive indicators were coded when the camera angle was from below, the camera distance was close, the people were active and eye contact was direct. The index was calculated using the Janis-Fadner Coefficient of Imbalance, a method to create one continuous-level variable while eliminating the likelihood of positive and negative emotions canceling one another out if the indicators were summed (Janis and Fadner, 1943).

Finally, physical touch was operationalized via an adaptation of the Goffman (1967) dual touch system of reciprocal and non-reciprocal. To keep in line with the aforementioned non-verbal indicators, touch was coded as negative (such as placing handcuffs on a subject), positive (such as a handshake), neutral (such as incidental contact), or no touch.

Data analysis

Both authors coded the data. After a codebook was developed, tested, revised and re-tested, a sample of 148 stories (12%) were selected for inter-coder reliability, which was calculated using ReCal (Freelon, 2013). All variables resulted in at least substantial agreement according to Landis and Koch (1977), ranging from 0.77 to 1 (see Table 2).

OPEN IN VIEWER

Table 2.

Intercoder reliability.

Variable	Reliability (κ)
Primary topic	0.79
Success factor	0.93–1 a
Caption mention success factor	0.86
Photo credit	0.86
Photo depict success factor	0.85
Photo dominance	0.9
Elbert’s hierarchy	0.89 b
Camera angle	0.83
Camera distance	0.77
Person(s) present	0.96
Activity in photo	0.9
Physical touch	0.95
Eye contact	0.89

a

Range reported due to multiple variables.

b

Krippendorf’s alpha used due to ordinal data.

Crosstabs and frequencies were used to analyze the data, and z-scores used to normalize data when necessary.

Topical factors

Several insights surrounding topical factors were discovered. Photos that reflect the success factors are likely to be feature photos, and photos that do not are likely to be spot news and portraits. This aligns with what we know about each type of photo, especially feature and spot news, and gives us insight to the use of portraits and solutions journalism. A feature photo is likened to something timeless, with human interest, and we could see how that was reflected in the population. Stories about people coming together to solve a problem are a better fit for a feature-type photo than a general news image. Spot news photos generally involve a conflict as it is happening, such as a protest, accident or crisis. We saw spot news photos especially in stories dealing with how communities are tackling crime. While the story described a solution, the included spot news photo instead showed a police officer making an arrest. In some cases, increased police presence was the solution, so the spot news photo could reflect the success factor, but this was seen rarely.

The use of portraits as the main image provided interesting insight. It was, at first, surprising to realize that a photo reflecting a success factor is less likely to be a portrait than a photo that does not, but this ties into an item on the Solutions Journalism Network’s list of ‘solutions imposters’. The list defines types of reporting that are close to solutions journalism, but not quite effective. One item is ‘hero worship’, which cautions writers to not focus on a sole person as the savior of the problem (Bansal and Martin, 2015). A portrait would place emphasis on a person and not on the solution. In some cases, however, such as a portrait of a worker who builds tunnels under roadways so animals can safely cross, the solution is depicted in the background.

By story topic, culture and education stories were the most likely to include a photo reflecting a success factor, while public safety and affairs stories were most likely to not have one. Again, we see a connection to the topic itself in how a story about a downtown restoration project that has added bright murals to old buildings could have a photo reflecting a success factor more easily than a story about improving the streetlights to increase safety. It’s interesting that environment and health stories had no clear leaning. Both topics conjure numerous visual depictions of success factors, yet environment was the topic with the lowest percentage of photos reflecting this.

Proximity factors

In our analysis, a proximity factor emerged concerning the photographer’s relationship to the media outlet. An internal photographer, such as a staff photographer, is most likely to take a photo that depicts a success factor. It is reasonable to believe the staff photographer may be better equipped with knowledge of the goals of the story and might even work directly with the writer. In some cases, the author and the photographer are the same person. As you move further away from the newsroom, such as to a wire service, the likelihood of such photos decreases. In these instances, the page editor is searching and selecting from pre-existing photos. It may be harder to find a particular photo that fits the goals of the story, and so the next best step is to choose an image that shows the conflict addressed by the solution. Even photos provided by outside sources, such as a freelancer or media relations staff, are less likely to reflect a success factor. One interesting exception to this is in the placement of the resulting photo – non-dominant photos, which are smaller and further away from the main parts of the page, were more likely to reflect a success factor. We see this as supporting how graphic photos (such as those that depict conflict) are more attention-getting to the reader and may therefore be elevated to dominant positions more often (Fahmy et al., 2014; Knobloch et al., 2003).

Human factors

Perhaps the most interesting insights come from human-related factors. When looking at success factors, the top four most likely to include photos reflecting them used wording specifically referring to humans or humanity: empowering people, embracing the power of relationships, practicing human centered design and cultivating collaborations (emphasis added). The bottom four did not. Of those top four, two were the only ones more likely to have a success factor depicted in included photos. We see this as an insight into the nature of humans and the news. When humans were specifically included, the chance of seeing photos reflecting a success factor increased. It’s a common understanding among photojournalists that imagery that includes people is more compelling, and we saw that reflected in the overall population of photos (82.4% had one or more person). While humans are involved in every one of the eight success factors, it appears this may connect with the extent to which they are involved.

Turning to the individual human, we discovered several key findings regarding emotion. Specifically, graphically appealing and emotionally appealing photographs reflected success factors more and informational photographs reflected success factors less. This tells us that when a photographer takes time and effort to move a photograph past the point of supplying basic information, it will increase the chance of reflecting the solution. We know that each level past informational causes a stronger response in the viewer (Kobré, 1999), so it is reassuring to proponents of solutions journalism to see that photos reflecting a success factor are found at the levels of impact.

At the latent, non-verbal level, while more negative indicators were seen overall and even in photos reflecting a success factor, we saw a small but significant association with the Janis-Fadner Coefficient of Imbalance. The more positive indicators, the more likely the photo reflected a success factor. This supports the Elbert Hierarchy in how these smaller indicators may be part of what makes a photo more than informational by adding emotional appeal. We also see that positivity is associated with solutions, which may sound inherent yet is an important finding. In regard to touch, we see similar results in how positive physical touch is more likely to be seen in a solutions photo than negative touch.