Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community for help and clear up your study doubts
Discover the best universities in your country according to Docsity users
Free resources
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
A study investigating the relationship between image fluency and individuals' perception of image veracity. The researchers manipulated image fluency using a photography filter and found that individuals' judgments of image truthfulness were influenced by the image's perceptual fluency. The study also revealed that disfluency, rather than fluency, had a stronger effect on the judgment of false images.
Typology: Study notes
1 / 53
This page cannot be seen from the preview
Don't miss anything!
RELATORE CORRELATORE CANDIDATO
In recent years, studies and experiments regarding heuristics and biases in judgments under uncertainty have proliferated. Nonetheless, few of these focused on visual communication, thus on information conveyed using symbols and imagery. Considering that this latter context can represent a fertile soil for heuristics and biases to arise, this study intended to explore these subjects in the visual context. Specifically, we tried to study the well-known Illusory Truth Effect in the visual context, that is, by using images as stimuli, instead of words. The classic Illusory Truth effect shows that sentences that are perceptually fluent are more likely to be judged as true. In our own study, we manipulated image fluency by using a photography filter (image exposure): darker images represented the disfluent condition, brighter images represented the fluent condition. The objective was twofold: 1) test whether image fluency affected individuals’ judgment on image veracity; 2) test if fluent images were more likely to be perceived as true. In order to accomplish this objective, we conducted an experiment in which each of 102 respondents saw 16 images, and then they were requested to judge their actual veracity. Respondents were randomly divided into two groups. Within each group, a random sequence of false/true, fluent/disfluent images were shown, such that images shown in the fluent condition in one group were in the disfluent condition in the other group. Thus, an ANOVA was run to test the two hypotheses. Our findings show that images’ perceptual fluency does indeed influence individuals’ perception of their actual veracity. However, fluent images are not always perceived as more true. What we do find is that inducing disfluency in picture viewing makes actual true images seem less true and actual false images seem less false. In our discussion we propose how this study can lead the way for an all-round exploration of heuristics and biases in visual communication. Moreover, we explain why this findings are relevant both in marketing and public policies contexts. Keywords: Illusory Truth Effect, visual communication, heuristics and biases, public policy tools.
Nowadays people are overwhelmed by a massive amount of information, advertising, and news, and most of it unsolicited. It comes from countless sources, an increasing proportion of it consists of visual information (pictures, video, …). Many observers of this phenomenon have argued that it has become impossible for people to process all the information they receive (Hasher, 1979 ). In such a context, it would simply be impossible for people to put enough effort into analysing, and therefore truly understanding, each message they receive. Thus, it happens that much information is passively assimilated and elaborated by the use of heuristics and, therefore, it could be misunderstood. The use of heuristics in judgment does not necessarily imply poor decision outcomes. It let us save energies and effectively deal with a massive amount of information. It can happen, nonetheless, that problems requiring a much more effortful analysis are solved with heuristics and, therefore, these lead to a poor decision outcome. This situation is potentially exploitable by firms. Past research studied a multitude of heuristics, both in theory and in practice, and the Illusory Truth Effect is part of such literature. The Illusory Truth Effect is, generally, the tendency to believe that a statement is true if it is easier to process, or if it elicits familiarity, regardless of its actual veracity. A lot of firms know how it does work and, consequently, they design some of their advertising strategies to exploit this effect; for instance, we can easily recall advertising statements such as “ number one in its category” written on products’ packaging or ads, these are intended to work through repetition, indeed, by reading again and again such statements, individuals will likely start to believe what they read. This is an example of the Illusory Truth Effect caused by repetition, and the resulting experience of familiarity with a given statement. The Truth Effect can work even in the short run. Indeed, by creating a statement that is easy to process, it is more likely it to be believed. As a result, these statements help to build favorable ideas of the brand/product and, therefore, improve brand image, likeability and so on. All prior studies on illusionary truth used text-based experiments and, therefore, they are valid only in such context. Even if texts and words constitute a significant
Firstly, images convey information as words do. But they do it in a different manner: by using a different level of communication. While textual communication takes places through argumentation, visual communication works by association (Muller, 2003). Therefore, an image is not true (e.g. fake) if it conveys information that is not real, or are not an accurate reflection of the reality, that is, it elicits associations that do not correspond to the reality depicted by the image. For example, this picture on Instagram (see Exhibit 1) conveys information to the social network’s users. Even if it is not explicitly stated, the image is visually communicating that these products resist and work well in such environmental situations; indeed, it displays Norrona’s products being used in an extreme natural environment. If the picture (Exhibit 1)^1 turns out to be unreal (i.e. photoshopped), this means that the image is not true, thus the associations which it elicits are not true. On the other hand, if an image conveys true information, and elicits associations that are coherent with reality, it is possible to define it as true. A common issue within images veracity lies in travel photographs and what experience they let you imagine. Among travel bloggers, using intelligent perspectives and Photoshop when making and modifying photos is becoming more and more popular. Their activity consists mainly of traveling around the world and post photo coverages on social networks so that their followers get inspiration. By using Photoshop, for example, their travel experience can be perceived as different from how it actually is (Exhibit 2). In Exhibit 2, it becomes clear how people’s perception of travel experience can be distorted from reality because of an image. Indeed, in bloggers’ photos, the Great Wall of China seems like a peaceful and wonderful place. But in the photo in Exhibit 2, it seems everything but a peaceful place. Of course, this matter is not limited to travel bloggers and Instagram. Potentially, every image can induce a misperception of reality because, like words, images are not reality itself, but only a representation of it. It is even critical to investigate, within this study, whether people are not as used to think critically about visual information as they are with textual information. If individuals will result to be less critical when they face visual information, thus being less capable to detect false information, this would result in a much more powerful Truth Effect in the visual context. (^1) Exhibits (Exhibit 1, 2, 3, 4, 5, and 6) are shown in the Appendix of the text.
Public Policy Tools Being welfare the basis of modern societies, it is clear why it is extremely important for democratic governments to effectively influence the behaviour of citizens. Public policies with the intent of influencing individuals’ behaviours, aim to improve their welfare and are, in general, in the interest of the public. To implement public policies, governments use public policy tools such as legislation, sanctions, regulations, taxes, and subsidies (Andrews et al, 1995). One tool governments can dispose of, according to the Office of Technology Assessment, is information reporting (Andrews et al, 1995). This specific tool aims to affect individuals behaviour mainly by delivering relevant information regarding a problematic issue, for example, texts such as “alcohol abuse can kill you” provide information which is meant to alter individuals behaviour. Chosen information usually refers to environmental concerns or individuals health issues. Visual communication is a relevant issue in public policies. Indeed, in recent years information reporting showed an increase in the use of pictures and graphics, which demonstrated to be a practical and effective tool for affecting people’s habits. This study will take into account one of the most known examples of visual communication as a public policy tool: warning labels on cigarettes’ packaging (see Exhibit 6).
As suggested by Gilbert (Gilbert, 1991, s.116), people are very much prone to accept as true things they hear and see, that is, they are truth biased. This happens because our brain faces both, time and energy restraints when thinking; therefore, in order to be efficient, it may fail in unaccepting the ideas it involuntary accept during comprehension. In addition, it is important to note that not all ideas are rationally analysed, instead, our brain actively analyses only a very small percentage of all retained information. Therefore, individuals usually end up believing false information. Even if Spinoza’s view seems to be reasonable, until recent years the “Descartes’ view”, regarding the mental process to assess veracity, was commonly and widely accepted. Only with the growth of scientific research in the field of heuristics and biases (about how our brain often works intuitively and automatically), Spinoza’s work started to be considered. Therefore, in order to obtain a deeper knowledge, the Cartesian account is outlined. Descartes argued that, in contrast with Spinoza, not all ideas were accepted but, instead, they were all comprehended, analysed and then accepted or rejected (Descartes, 1641/1993). “Figure 2” shows how the two different views operate with false information. Mmmmmm Figure 2. Source: Descartes Versus Spinoza: Truth, Uncertainty and Bias. The main difference is that in the Cartesian’s view our mind is able to comprehend an idea without assessing its veracity, and there is an initial period of non-decision and then the evaluation (Street & Richardson, 2015). On the other hand, Spinoza explained that an automatic belief always takes place and, then, a cognitively effortful phase of evaluation follows.
Gilbert (Gilbert, 1991, s.108) provided a smart example to clarify how the two alternative views work and differ from each other: the librarian example. Imagine the human brain to be a library, and the information gathered to be several volumes. There are at least two ways by which one could tag the spines of books in order to distinguish which one is fiction and which is not (false and true information gathered). One possibility consists in reading all the books and determine whether they are fiction or not and, therefore, to tag the volumes by using two different colours, depending on the books’ nature. This process’ major advantage is that the librarian will be sure about which book is fiction and which is not, but its disadvantage lies in the fact that the process requires much effort and time. This is the Cartesian account. On the other hand, as Spinoza argued, the alternative method is to use a tagged- untagged system, in which tagged books are fiction, and untagged ones are non- fiction. In Spinoza’s view books were represented before being assessed but, because of its particular system, a new volume without a tag looked exactly like a non-fiction book, thus like true information. The new volumes will be placed in the shelf without any tag, and just in case the librarian finds out that the book is a fiction work, he will tag its spine. Here, the advantage is that the librarian doesn’t necessarily have to read all the books before placing them on the shelf, thus, it is possible to deal with a huge amount of information. Instead, the obvious disadvantage consists in the fact that the librarian, before he individually reads the books, may tag fiction books (false information) as nonfiction books (true information). Finally, even if the Cartesian account had never been really contested until recent studies regarding automatic psychological processes, the evidence in support of Spinoza’s view is growing nowadays; therefore, it confirms that individuals could be often truth-biased in their judgments. Fluency Causes the Illusory Truth Effect Spinoza argued that our brain assumes most of the information it receives as true. But it was still not clear how this process took place. The philosopher understood that individuals are not able to deal with the huge amount of data he faces, therefore, on the base of this intuition, he built its theory. Modern psychologists found out that this “truth assumption” happens when our brain is working with low effort, that is,
Visual Communication vs Textual Communication Even if available research about visual communication is far narrower compared to textual communication, the former is deeply inherent in human nature, while the latter has to be taught to individuals by individuals. Images are not only easier to recognize and process compared to words, but they are also less effortful to recall. When words enter long-term memory they do so with a single code (textual code). Images, instead, contain two codes: visual and verbal, and each one is stored in different places in the brain (Paivio, 1971). Visual communication is, basically, the conveyance of information and/or ideas through the vision. It occurs through different kind of visual content, such as images, videos, body language, etc. (Müller, 2007). An interesting difference in comparison with textual communication is that “images are not read, they are seen” (Müller, 2003). This distinction is relevant, and it suggests that different mental processes are used when individuals face texts and images. Even if both, textual and visual communication convey information, they do that in a different way, through different instruments and channels. It is not possible to affirm which one is more effective, in the same way as it is not possible to tell whether it is better to travel by train or by airplane (otherwise they would not coexist). Probably, these two ways of communicating have more similarities rather than differences. For example, both of them can lead to illusions of truth, that is, both of them can convey wrong or false information. The well-known Muller-Lyer illusion, showing two lines, which appear to be different in length, but they are not, is a simple example of an illusion of truth in the visual context (see Figure 3). Figure 3. Muller-Lyer Illusion.
Visual Communication as a Policy Tool In this study, we focus specifically on warning messages on cigarettes’ packaging as a public policy tool, because of the extensive available research regarding graphic warnings on tobacco products. Nowadays, almost everyone has seen the scary pictures printed on cigarettes’ packs. They represent various health diseases caused by smoking tobacco. They are chosen to have a strong emotional impact on individuals, specifically, they tend to evoke a feeling of fear (Kees, Burton, Andrews, & Kozup, 2010). Examined past research has proven that visual communication, in the form of graphic warnings, is effective in influencing individuals. For example, it has been demonstrated that pictorial warnings are significantly more effective than text-only messages (Hammond, 2011). Moreover, the graphic warnings evoke fear, in turn, fear mediates the effect graphics has on the intention to quit smoking (Kees, Burton, Andrews, & Kozup, 2010). Finally, there is sufficient evidence suggesting that this form of visual communication increases the knowledge of risk and could lead to smoking cessation and prevention. Other studies confirmed what Hammond wrote, for example, O’Hegarty et al showed that text-plus-graphic warning labels are more salient than text labels, this lead to the conclusion that visual communication is an effective and necessary public policy tool. Scary pictures on packs effectively influence individuals because they firstly comprehend the message behind the photos and, secondly, find it relevant for themselves. Therefore, it is crucial whether these people judge pictures as real or not, that is, whether they consider them depicting real health diseases caused by smoking or unrelated diseases. In case individuals believe images on warning labels as false or not related to smoking, there would be no consequent intention to quit smoking. As explained before, judged veracity of images is related to the actual underlying message, therefore, people thinking a warning label to be false means they think the message conveyed by the label is not true, that is, that disease is not caused by smoking or the image is fake. Finally, the Illusory Truth Effect can be a relevant issue for graphic warning labels specifically, and public policies tools (such as information reporting) in general.
image in Exhibit 3.1 is true or false is equivalent to judge about the veracity of the moon landing. Experiment The experiment is an adaptation of the one conducted by Reber and Schwarz (Reber & Schwarz, 1999), with which it had been tested the illusory truth in textual communication. The experiment consisted in showing several sentences to respondents, both true and false, manipulated to be more or less fluent, thus easy or effortful to read. This was made possible, in short, through the use of blurred filters to create the disfluent condition and sharp and clear font for the fluent condition. In the current experiment, one hundred two undergraduates and graduated people between 18 – 25 years old answered the survey. Sixteen images were presented to each respondent. Half of these images were true, and the other half false. Perceptual fluency was manipulated by altering the image light exposure, that is, rendering images dark or bright. Easy to process images were brighter; hard to process images were darker. Therefore, sixteen selected images were shown to respondents, and the manipulation of fluency occurred between groups, thus each image appeared brighter to half the respondents and darker to the other half. The random sequence of true/false and fluent/disfluent images is described later in the text and observable in the following pages (see Figure 3.1 to 3.16)^2. Respondents were asked to judge on a scale from one to seven whether they agreed the shown image was true. With “1” being “ completely agree ”, and “7” “ completely disagree ”. A brief but clear explanation about what the researcher meant with true and false images was given to participants. Also, in order not to bias respondents’ judgment because of the manipulation (dark/bright), participants were told that some images were dark to calculate the difference in response time to different stimuli. Finally, demographic and education questions were asked at the end of the survey. The whole survey required an average of three minutes time for respondents to (^2) The sequence is shown from page 14 to page 16. Please note that due to the reduction in size and lower image quality in this document, dark images seem to be too dark and incomprehensible. In the study they appeared to respondents on mobile devices in bigger size and higher quality.
complete it. The experiment was made short with the purpose not to annoy participants and therefore alter their judgment quality. Qualtrics online software was used to realize the survey. The questionnaire design consisted of a first description of the experiment to respondents; then, respondents were randomly assigned to one of the two groups. The two groups basically saw the same images, but the images that were bright for the first group were dark for the second group, and vice versa. The final demographic and education questions were the same for all respondents. Exhibit 4 and Exhibit 5 show further information and provide examples of the questionnaire presented to respondents. Selected Images All images are taken from the web and the majority of false images are photoshopped. They show different subjects and they are relevant in different contexts. The aim was to find contexts in which images play a key role in influencing people’s behaviour. Indeed, by looking at Figures from 3.1 to 3.16, it appears that Figure 3.1 and Figure 3.8 are relevant for the fashion industry; Figures 3.2, 3.4, 3.6, 3.11, 3.13, and 3. 15 are relevant within the political environment; Figures 3.3, 3.7, 3.9, and 3.16 focus on branding, product placement and advertising in general; Figures 3.5, 3.10, 3.12, and 3.14 are relevant in the food industry. Within the eight true images, five images are easier to recognize as true (Figures 3.1, 3.3, 3.9, 3.10, and 3.14), and three images are more difficult to recognize as true (Figures 3.11, 3.13, and 3.16) because they convey information that is unusual or unlikely to be true. On the other hand, four false images are easier to recognize as false (Figures 3.4, 3.7, 3,12, and 3.15), instead, the other half describes a more likely situation (Figures 3.2, 3.5, 3.8, and 3.16). False photos are divided into two categories:
Figure 3.7: Alpinist on the mount Everest with a very light gear. Figure 3.8: an unnatural female model posing. Figure 3.9: a random photo from Starbucks. Figure 3.10: a random food photo from Instagram. Figure 3.11: Us Marines soldiers under the Fuji mountain. Figure 3.12: a perfect looking fast food’s hamburger.
Figure 3.13: Kim Jong Un and Donald Trump handshake. Figure 3.14: A random food photo from the web. Figure 3.15: Matteo Salvini in the role of a immigrants’ teacher. Figure 3.16: a dreamlike resort. Data Collection and Preparation The data have been gathered with the use of online surveys powered by Qualtrics. The survey was spread via a link, mainly on mobile devices such as phones. There were no privacy concerns, indeed, the only personal information asked were basic demographics and the whole questionnaire was anonymous. Gathered data were analysed with statistical software (SPSS) in order to detect any significant pattern in the results and test the hypotheses. Nonetheless, data gathered through Qualtrics were not yet ready for running statistical analyses. For this reason, a consistent time has been invested in data preparation in order to obtain a dataset that fitted the necessary statistical analyses on SPSS. In particular, the original
