Vivid visuals from far-flung conflicts regularly occupy our news screens. The coverage of Islamic State in Iraq and Syria is a prime example. Although geographically distant, images that are beamed across the globe can provide us with an immediate emotional connection: disgust and anger at picture of a to-be-beheaded journalist, fear at a masked militant, or sympathy for the millions of suffering refugees that lie in their wake. What is less clear, however, is whether emotionally charged news visuals can influence public opinion toward substantive policies. Indeed, scholarly knowledge of the role of images in media effects lags behind today’s increasingly multimodal media landscape.
By Menendj (File:Iraqi insurgents with guns.JPG) [CC BY-SA 2.5 (http://creativecommons.org/licenses/by-sa/2.5)], via Wikimedia Commons
To address this issue, my supervisors and I devised an experiment using image and text exemplars of two well-trodden news frames from international affairs. Specifically, this included the frame emphasising an obligation to intervene to help suffering victims of a conflict, versus the frame outlining the risks of intervening posed by dangerous militants. These counter-frames have been prominent since the 90s during US incursions in Somalia and the Balkans, and more recently in coverage of conflicts in Libya, Syria and Iraq.
In our study, we used news material about a lesser-known conflict in the Central African Republic (CAR) to avoid prior knowledge biasing results. We constructed several online news articles containing an image either of victims or militants from the conflict, alongside a news story with a few phrases modified to present the obligation or risk frames. We then combined these images and texts in congruent and incongruent pairs (i.e., matching or non-matching image and text), giving the experimental conditions shown in Figure 1.
Figure 1. Experimental conditions: Stimulus articles were comprised of congruent and incongruent combinations of a news image and text.
After reading one of these articles, participants answered questions about their support for military intervention in the CAR, and their intention to discuss, donate to a cause, sign a petition and join a protest in support of those embroiled in the conflict. Finally, participants were asked to indicate how much sympathy, fear and anger they felt when viewing the article.
We expected that congruent image-text pairs – endowed with the salience and vividness of visuals plus the guided structuring of text – would have the strongest effect on support for intervention. When incongruent, prior research suggested that the frame conveyed by attention-grabbing images may dominate and override divergent textual information.
In fact, our results showed that congruence was not the decisive factor. Instead, opinions about support for intervention were determined by the framed text regardless of the accompanying image. In contrast, behavioural intentions (to donate, sign a petition, etc.) were driven by the frame of the article’s image irrespective of the paired text.
Going further, strong emotional responses to the images helped explain these effects. Participants’ feelings of sympathy to the images of victims predicted their behavioural intentions. In contrast, emotional responses did not explain the changes in participants’ opinions, which were determined by the frame of the text. So, emotional responses were particularly important when images influenced our participants’ behaviour, but less so when text guided them to form an opinion.
These findings go beyond previous studies to show that frames presented in images and text can play a unique role in influencing an audience. Moreover, it seems that the emotional consequences of images can play a formative role in generating an interventionist public response.
What are the implications of these findings for democratic policy-making that reflects public opinion?
Our results chime with previous studies showing that vivid visuals can help typically apathetic citizens to make political decisions through minimally effortful heuristics (or rules of thumb). This can be positive when the situation demands an active response, such as charity appeals for natural disasters, or for raising a grave but little known conflict up the public agenda. However, deciding complex matters of foreign policy solely on an empathic emotional response is not uniformly desirable. An angry response to the beheading of western journalists, for instance, should not be the single determinant of an intervention against Islamic State. Fortunately, our results suggest that for opinion formation people are receptive to nuances outlined in an article’s text.
Ultimately, normative conclusions about multimodal media effects depend not only on characteristics of the message, but also on the audience and media outlet. The interaction between these factors is an important avenue for future research, and should provide a clearer picture to those in the newsroom about the role of visuals in media effects.