Table 1 - uploaded by Adrian Furnham
Content may be subject to copyright.
Source publication
This study looks at how people rate specific non-verbal cues (NVC) with regard to the presence and persuasiveness of the speaker. The first study (N = 1500) was conducted to attempt to isolate specific non-verbal signals (hands and feet) that were related to ratings on influence and persua-siveness. In the second study, over 600 people rated a high...
Citations
... Sec. 2.3.1) as it states that gestures, gaze, and hand movements are significant indicators of perceived persuasiveness (Newman et al., 2016;J. Peters & Hoetjes, 2017). ...
... This refers to the correctness or faithfulness of the generated explanation (Huber et al., 2022;Mohseni et al., 2021). Within the context of this work, we argue that our network can be considered faithful because its focus aligns with findings from existing literature (e.g., Newman et al., 2016;J. Peters and Hoetjes, 2017), making it reasonable to assume that the explanations are correct and valid. ...
... Using common XAI methods, we are able to generate explanations demonstrating a network's focus on persuasive cues, such as the speaker's contours, gestures, and hands. This aligns with known persuasive indicators from the existing literature (Newman et al., 2016;J. Peters & Hoetjes, 2017), proving the validity and fidelity of the generated explanations. ...
With the steadily growing abundance of online information, whether through news portals on the internet or social networks, two increasingly pressing problems have arisen: On the one hand, online users in social networks are frequently confronted with distorted and one-sided information due to filter algorithms, while on the other hand, there is a diminishing willingness for open discourse.
These issues have been particularly evident during the recent COVID-19 pandemic, where divergent viewpoints were quickly rejected. One reason is that it is easier to engage with arguments from one's own side, while arguments from the opposing side (referred to challenger arguments) are often blocked or perceived as provocative.
This phenomenon can be explained psychologically through affective reactions and peripheral information processing, where information contradicting one's own opinion triggers strong emotional reactions, thus complicating rational understanding and leading to a reflection bias. This means that certain arguments are either misinterpreted or ignored (content-based reflection bias), or people are influenced by subliminal cues, such as emotions (behavior-based reflection bias), of which many people are not aware, though.
This dissertation focuses on the reflection bias from two perspectives: 1) Raising awareness of the behavior-based reflection bias through the use of explainable Artificial Intelligence, and 2) mitigation of the content-based reflection bias using an argumentative dialog system.
In the first part of the thesis, we examine how, with the help of explainable Artificial Intelligence and Neural Networks, we can make the behavior-based reflection bias, specifically regarding gestures, visible. The goal is to draw attention and awareness to the influence of gestures on the perceived persuasive effect through visual explanations.
We investigate whether we can generate satisfactory explanations when training Neural Networks with subjective data that significantly differ in quality and accuracy from gold standard data due to noise. Furthermore, we explore whether these explanations are suitable for highlighting the behavior-based reflection bias and differences between individuals by examining whether the focus of the networks aligns with insights from the literature.
Our analysis shows that Neural Networks primarily focus on hand gestures, which is identified in the literature as an important indicator of persuasion.
In the second part of the thesis, we explore how to mitigate the content-based reflection bias. To achieve this, we develop an argumentative dialog system that encourages users through targeted interventions to move away from a one-sided argument exploration and to consider arguments from the opposing side. The system utilizes a metric, gauging the extent to which users predominantly focus on arguments that align with their viewpoint. In three studies, we examine the effects of interventions on reflection and exploration behavior.
The results of the studies demonstrate that users significantly engage more with challenger arguments and spend considerably more time considering these arguments when the system applies intervention strategies. Additionally, we present some interaction effects with personality traits, supporting the idea that systems aiming to improve reflection should also take into account the user's personality.
... Non-verbal communication cues can influence the message's perception and processing (Newman et al., 2016;King et al., 2020). We use our body and voice changes to reinforce or qualify what we are saying, convey emotions, attitudes, and intentions, regulate the flow of communication, establish contact, and provide feedback with/to other people. ...
Some non-verbal cues, such as voice pitch and gestures, can influence the individual's emotional response to different messages. For this reason, non-verbal communication can be understood as the language of emotion, while the content is the language of ideas. But the research question here is what voice pitch and hand gesture strategies are more effective, attractive, and elicit a stronger emotional response. To answer this question, this study analyzes some pitch and hand gesture strategies in public discourses. In the experiment, 120 participants were exposed to different public presentations with three conditions for voice pitch (few, medium, and many variations) and three for hand gestures (smooth, medium, and strong intensity). Then, they rated the effectiveness and attractiveness levels and self-report arousal and valence while their emotional response was registered. Physiological arousal was measured with electrodermal activity (EDA) and valence with facial expression recognition. Participants perceived the medium pitch variations and the medium intensity of gestures as the most effective and attractive strategies with the highest self-reported arousal and valence. The combination that elicited the strongest autonomic arousal was medium pitch variations-strong gestures. Finally, the most positive emotional valence was for many pitch variations and medium gestures and the happiest for the combination many-strong.
... GIFs were chosen as they merge the image format of old media with new media (Dobson & Knezevic, 2018) through short moving pictures (Kanai, 2016). A central affordance of GIFs as memes are their ability to convey dynamic facial emotion (Kanai, 2016;Miltner & Highfield, 2017;Newman et al., 2016). It is not surprising that GIFs are often used in social media to express emotion (Gürsimsek, 2016;Miltner & Highfield, 2017;Newman et al., 2016) and emotion within text-based communications (Kanai, 2016;Miltner & Highfield, 2017). ...
... A central affordance of GIFs as memes are their ability to convey dynamic facial emotion (Kanai, 2016;Miltner & Highfield, 2017;Newman et al., 2016). It is not surprising that GIFs are often used in social media to express emotion (Gürsimsek, 2016;Miltner & Highfield, 2017;Newman et al., 2016) and emotion within text-based communications (Kanai, 2016;Miltner & Highfield, 2017). GIFs are examples of popular and ubiquitous multimodal formats of message propagation through social media (Knobel & Lankshear, 2007;Shifman, 2013). ...
... Competence posture Newman et al. (2016) and Rennung et al. (2016) noted that persons with an open and erect posture were perceived as more competent than those in contracted postures. As seen in Appendix B, open and erect postures signal dominance and power, a feature related to competence (Kervyn et al., 2015). ...
Memes within animated graphical interchange formats (GIFs) are developed and shared by Internet users to communicate cultural ideas, symbols, or practices for a wide global audience. Among the billions of GIFs shared internationally, some portray scientists engaged in scientific work. Media and science education scholarship alike have evidenced how scientists are portrayed can influence social perceptions of science and contribute to stereotypes that deter youth’s interest in and affinity to science and science occupations. To understand what social perceptions of science may manifest from new media (GIFs), the present study ascertained stereotypes using Warmth and Competence constructs from Fiske’s Stereotype Content Model (SCM). The SCM utilizes high, medium, and low warmth and competence dimensions found in media-based imagery to illuminate stereotypes. Researchers coded and categorised 287 meme-based GIFs of scientists sourced the largest online GIF repository, Giphy. A directed qualitative content analysis found high-competence and low-warmth dimensions most represented within the sample that theoretically (per SCM) represent perceptions that contribute to an envious stereotype with elements of admiration and contempt. This study suggests that although there have been improvements in the portrayals of scientists in media, however, GIFs may preserve and perpetuate the trope of the competent, yet cold, scientist.
... Thus, if the feelings conveyed by the facial expressions or body language of individual A tend to be imitated passively or subconsciously by B, the perception of the imitation by A creates rapport with B (Chartrand and Bargh 1999). Sundaram and Webster (2000) proposed a typology of NVIC that stands out due to its emphasis on the impact that it may have on the satisfaction of customers when practiced by FLSEs: a) paralanguage: tone, intensity, vocal amplitude, fluency or pauses, used during transmission of a message as hints to identify the real emotions of the speaker (Sidtis and Kreiman 2012); b) kinetics: body movements, posture, eye contact, head inclination, hand movements and facial expressions (Newman et al. 2016); c) personal appearance of the service provider: personal hygiene, clothing style, state of conservation of uniforms (Sundaram and Webster 2000); and d) proxemic communication: nonverbal communication that involves the unconscious structure of people in space, such as the distance between people when engaging in daily transactions (Hall 1963). ...
In this paper we use qualitative data to identify how pharmacy frontline sales employees try to build rapport toward embarrassed clients using nonverbal interpersonal communication (NVIC). Upon doing so, we build on a conceptual framework where NIVC can create rapport in embarrassing situations, with the client usually preferring the salesperson who, through expressive similarity, connects himself/herself with the client’s emotion and behaves discreetly, preserving the client from unnecessary exposure. Four methods were triangulated, including the critical incident technique. Results show that rapport building on the part of these salespeople by means of NIVC toward embarrassed clients is eminently amateurish, as these employees do not receive any prior training to perform in a more professional way. A relevant academic contribution comes from challenging the preconception that the categories of rapport are mutually exclusive and exhaustive, allowing the initial conceptual framework to be reconfigured as a new category of rapport is included. In practical terms, this research is also distinctive, since it suggests that, as a peculiar locus for embarrassment, pharmacies take these clients’ emotion in a more serious way, searching for ways to mitigate it through better training of salespeople, as this effort may translate into higher customer satisfaction and loyalty.
... However, there is still no empirical confirmation for a universal hierarchy of effectiveness, attributing stronger effects to nonverbal cues, and debates continue about whether verbal or non-verbal cues in speeches have greater influence on the impression gained by the audience. What seems important is that speakers orchestrates consistency in the overall set of cues they convey (e.g., Isbister & Nass, 2000) and following simple rules such as standing still, using a lower pitch, standing with feet apart and employing congruent gestures help speakers to be rated as confident and persuasive (Newman et al., 2016). Research suggests, however, that such non-verbal features may only temper an audience response rather than sway it (Nagel et al., 2012;Jackob et al., 2016). ...
Stance, the extent writers intervene in a text to convey their personal attitudes and assessments, has long been a topic of interest to researchers of academic communication. Less studied, however, is how stance functions in spoken discourse. This would seem to be a particularly important issue in the Three Minute Thesis presentation (3MT), a relatively new genre which captures the competitive and high pressure atmosphere of the modern academy. In these competitions doctoral students present their research using only one static slide in just 180 seconds to a panel of judges and non-specialist audience. Using Hyland's (2005) model, we explore speakers’ interactional and evaluative positions in a corpus of 140 3MT presentations from the physical and social sciences. Our findings show that this monologic genre is heavily stance laden and that speakers from the hard and social sciences adopt different stance positions. Hard science students take a stance by casting doubt or asserting certainty in the reliability of information while social scientists claim an authorial self through a more visible personal presence and explicit affective commentary. Our findings have important implications for understanding academic speech genres and for EAP teachers preparing students to orally present their research.
... Schmid Mast and Hall (2004) found that participants could accurately guess colleagues' relative positions in an occupational hierarchy based on their body positions, broadly defined, and participants are more persuaded by a speaker who engages in expansive body postures compared to contracted postures (McGinley et al. 1975). Participants rate people adopting expansive postures as more dominant, convincing, competent, admirable, and a better leader than those in contracted postures (Newman et al. 2016;Rennung et al. 2016). Even in implicitless controlled-responding, expansive and contracted postures, respectively, activate impressions of dominance and submissiveness (Bailey and Kelly 2015). ...
... The evidence, however, that women's expansive body postures override gender stereotypes, without eliciting negative reactions, is mixed. Recent research using more explicit, controlled perceptions of power finds that men and women in identical body postures are judged similarly on power ratings (e.g., leadership; Newman et al. 2016;Rennung et al. 2016;Shariff 2010), but Henley and Harmon (1985) found that men in expansive body postures conveyed dominance to a greater degree than women. Bailey and Kelly (2015) also found evidence that gender interfered with power information conveyed by body posture, and men in contracted body postures did not implicitly activate low power concepts to the same degree as women. ...
... Participants completed the posture classification task from Study 2 as well as explicitly judging the leadership potential of presented targets. We expected that participants would explicitly rate individuals in expansive postures higher in leadership potential than those in contracted postures, regardless of target gender (Newman et al. 2016), while still displaying the implicit gender interference effects found in Study 2. ...
People rapidly form impressions of others based on their gender. Women tend to be liked more than men but men tend to be regarded as more powerful. However, a person’s nonverbal behavior has the potential to confirm or override these stereotypical impressions. Specifically, expansive, open body postures (e.g., based on widespread limbs) tend to convey high power compared to contracted, closed body positions. In three studies, we tested whether postural variations affected evaluations of men and women and impressions of their power. In Study 1, images of women elicited a more positive reactions than images of men in an affective misattribution procedure, but only when women enacted contracted body postures. In Studies 2 and 3, participants were slower to classify images of expansive postures as high power when enacted by women and slower to classify contracted postures as low power when enacted by men, but rated men and women similarly in explicit power-related judgments. Expansive body postures thus appeared to eliminate the usual positive reaction to women relative to men, but women still did not implicitly convey power to the same degree as men. Gender did not interfere with explicit, more controlled judgments of power. Together these studies demonstrate that gender implicitly interferes with perceptions of a person’s power, even in the presence of potentially individuating body postures.
... Nonverbal cues have many variety of usage. According to Newman et al. (2016) with a slight change of nonverbal cues can immediately change the atmosphere of the presentation. ...
... According to Newman et al. (2016) a slight change of nonverbal cues, can immediately change the atmosphere of the presentation as well as audience engagement towards the presentation. Burgoon et al. (2016) reported that nonverbal communication helps an individual to send a message across without even using any verbal communication and it is much more effective in conveying emotion and feelings. ...
... Decoding information utilizes knowledge one may have of certain received sensations. As stated by Newman et al. (2016) people become verbal communicators, they begin to look at facial expressions, vocal tones, and other nonverbal elements more subconsciously. After mastering these nonverbal cues, people will be able to decode nonverbal cues with ease and are able to identify what are the audience feeling during the oral presentation and encode a suitable respond to the audience to maintain that attention and keep the audience engaged to the oral presentation. ...
Oral presentations have become part of many course assessments. During presentations, presenters use verbal and non-verbal cues. Learners Nonverbal cues are used either directly or indirectly by presenters to express their feelings towards the oral presentation. Some common nonverbal cues often used are kinesics and vocalic. Kinesics is derived from the word "kinesics" which means "movement" and refer to the hands, legs, body and facial movements. Vocalic is the study of voice and it is used to address a message for example like the tone of voice This study is done to explore the influence of kinesics and vocalic in ESL orang presentation among ESL learners. 30 undergraduates responded to a questionnaire about how they perceive the use of kinesics and vocalics in oral presentations. Findings revealed ESL learners used a variety of nonverbal cues in their presentation for different reasons in their presentations. The results of this study bear interesting pedagogical implications for classroom teaching and learning of ESL oral presentations.
When interacting with sophisticated digital technologies, people often fall back on the same interaction scripts they apply to the communication with other humans-especially if the technology in question provides strong anthropomorphic cues (e.g., a human-like embodiment). Accordingly, research indicates that observers tend to interpret the body language of social robots in the same way as they would with another human being. Backed by initial evidence, we assumed that a humanoid robot will be considered as more dominant and competent, but also as more eerie and threatening once it strikes a so-called power pose. Moreover, we pursued the research question whether these effects might be accentuated by the robot's body size. To this end, the current study presented 204 participants with pictures of the robot NAO in different poses (expansive vs. constrictive), while also manipulating its height (child-sized vs. adult-sized). Our results show that NAO's posture indeed exerted strong effects on perceptions of dominance and competence. Conversely, participants' threat and eeriness ratings remained statistically independent of the robot's depicted body language. Further, we found that the machine's size did not affect any of the measured interpersonal perceptions in a notable way. The study findings are discussed considering limitations and future research directions.
