Dual-process theories of reasoning (e.g., Gilbert, 1991; Stanovich & West, 1997) posit that decisions are mediated by two cognitive systems: a fast and automatic system which sometimes relies on past knowledge, and a conscious and effortful system which is more likely to adhere to the rules of logic. Dual-process accounts of memory (e.g., Joordens & Hockley, 2000) suggest that memory is influenced by two cognitive systems: a fast and automatic familiarity component, and a conscious and effortful recollection component. Both accounts suggest that cognition relies on two underlying systems, which are described similarly in the two literatures, suggesting some form of convergence in these two areas of research. Memory research may therefore be informed by considering decision making research, and vice versa. Combining these two theoretical perspectives, it follows that believable evidence should be less memorable than unbelievable evidence due to its shallow initial processing. Despite this fact however, when inferences are being made based on evidence retrieved from memory, believable evidence should actually have a larger impact than it does when it is provided online, whereas no change or a lesser impact should be noted for unbelievable evidence. Across 3 experiments these predictions are validated, suggesting that the impact of evidence on inferences depends not only on the believability of that evidence, but also on whether the decision is being made online or from memory. Specifically, memory-based inferences exaggerate the influence of believable but not unbelievable evidence, despite the fact that unbelievable evidence is more memorable.

Dual-process theories of reasoning (e.g., Gilbert, 1991; Stanovich & West, 1997) posit that decisions are mediated by two cognitive systems: a fast and automatic system which sometimes relies on past knowledge, and a conscious and effortful system which is more likely to adhere to the rules of logic. Dual-process accounts of memory (e.g., Joordens & Hockley, 2000) suggest that memory is influenced by two cognitive systems: a fast and automatic familiarity component, and a conscious and effortful recollection component. Both accounts suggest that cognition relies on two underlying systems, which are described similarly in the two literatures, suggesting some form of convergence in these two areas of research. Memory research may therefore be informed by considering decision making research, and vice versa. Combining these two theoretical perspectives, it follows that believable evidence should be less memorable than unbelievable evidence due to its shallow initial processing. Despite this fact however, when inferences are being made based on evidence retrieved from memory, believable evidence should actually have a larger impact than it does when it is provided online, whereas no change or a lesser impact should be noted for unbelievable evidence. Across 3 experiments these predictions are validated, suggesting that the impact of evidence on inferences depends not only on the believability of that evidence, but also on whether the decision is being made online or from memory. Specifically, memory-based inferences exaggerate the influence of believable but not unbelievable evidence, despite the fact that unbelievable evidence is more memorable.

Non-Player Character (NPC) believability is a game aspect that can be exploited to increase a game's replayability, but little research has been conducted on the topic. One method for enhancing a NPC's believability is to integrate human-like behaviours into their design, so that they react to players in a realistic and interesting way. A large part of human behaviour can be explained by their emotions; therefore it was selected as the inspiration for the GLaDOS system. Two psychological theories of emotion, Lazarus's cognitive appraisal and Plutchik's psycho-evolutionary synthesis, guided the design of the GLaDOS system, although several components are not unique to these theories. An implementation of the design was created as a "mod" for the popular CRPG "The Elder Scrolls V: Skyrim" to test its feasibility within the context of a commercial game. This task required an additional psychological model, PAD space, to map appraisal values to emotion codes and intensities. Feasibility testing was done via a user study to determine if the GLaDOS system increases player engagement when compared to the original game. While the objective analysis found that there were no significant differences between the two versions, subjective participant responses expressed a strong affinity for the GLaDOS system. Since player engagement is inherently subjective, it is encouraging to see positive responses from participants. This indicates that the GLaDOS system, and NPC believability in general, is one aspect of video games that has the potential to increase a game's replayability and should be investigated further. / Thesis / Master of Applied Science (MASc) / Realistic video game characters are a desirable game component to increase a game's value. Even if the game's ending does not change, realistic character behaviours encourage players to replay a game multiple times to see what happens along the way. This is closer to tabletop games where players know the game's outcome, but still play because no two sessions are alike. Despite its advantages, few developments have been made towards realistic game characters. An easily recognizable factor of human decision-making and behaviour is emotion and integrating emotion into character design is one way to improve their realism. The GLaDOS system is a proof-of-concept product that incorporates psychological models of emotion into its design. To test its impact on player engagement, the system was implemented as an extension for the popular computer game "The Elder Scrolls V: Skyrim". Preliminary test results are promising and show that further development could prove fruitful.

This thesis focuses on the importance of technology to create believably composited effects. It was found that many factors culminate in generating believability in a film, including: suspension of disbelief, the story, and the quality of the special effects. Many technical aspects lend to the creation of successful special effects and are involved during every stage of production. There is a discussion of several of the important criteria analyzed during preproduction, production, and post production. A brief history of the technical effect industry is discussed. Personal work for this project includes three case studies. In the form of short video projects, these studies are applications of the researched industry concepts. They deal with issues including incorporation of digital models into live action footage, using pre-existing footage, digital makeup, motion tracking, masking, color correction, and generation of artificial lights and shadows. The creation of these videos included video recording and editing and used Maya TM and After Effects TM. The final shorts showed examples of the strengths and weaknesses of the applied compositing techniques. Implications for the future directions of this field are also discussed.

An essential part of the video game experience are the characters. How the characters are perceived as believable can make or break the immersion of the experience. This paper aims to create a model which can be used as a tool for analysis, designed to measure believability in video game characters. For consistency, we first define the term believability. Using thematic analysis, we aim to evaluate what components are relevant in the design process of a new model. This is achieved by studying three characters from japanese role playing games using two existing models to measure believability in artificial agents, as this hasn't been done in this context in the past. The results show that relevant components in existing models does indeed exist, but they lack some components mirroring feelings and personal development in characters from japanese role playing games. In addition to this, problems arising from the categorization of characters in models designed to measure believability is discovered. Two versions of the new model are presented, one designed to measure characters from various genres, and one designed to specifically measure characters from japanese role playing games. A different approach to categorize video game characters is also presented in the latter version of the new model.

Trovärdighet inom fiktiva världar är något som många strävar efter, men hur kan vi uppnå den känslan?  Inom film och litteratur används begreppet verisimilitude för att definiera ett verks sken av sanning eller verklighet och därmed även dess förmåga att framstå som trovärdig. Härstammat från Platons och Aristoteles mimesis, det vill säga imitation av naturen, utvecklades termen verisimilitude för att ge fiktiva verk möjligheten att undersöka och pröva hur de kan bli mer trovärdiga. För att nå trovärdighet pekar begreppet på hur väl åskådaren kan relatera verket till verkligheten, för om detta förekommer, kan verket genast bli mer trovärdigt.  Begreppet verisimilitude används främst inom fiktiva verk för film och litteratur, men det finns väldigt lite forskning om detta inom spelmediet. Vi anser att verisimilitude kan ha värde inom denna medieform för att ge speldesigners en bättre uppfattning om hur de kan designa spel mot trovärdighet. Med hjälp av Anneli Lehtisalo och hennes studier av filmen Runon kuningas ja muuttolintu har vi byggt upp en förståelse över verisimilitude och dess bruk inom filmmediet. Vi har även med hjälp av Kamal Bhatt och hans text Believability in computer games (2007) undersökt hur trovärdighet inom spel urskiljer sig från andra medieformer och vilka kvalitéer ett spel behöver för att vara trovärdiga. Med dessa förståelser har vi sedan i en gestaltning designat en fiktiv spelmiljö med verisimilitude som ett designperspektiv. Med detta menar vi bidra med en grund för hur en kan designa spelmiljöer mot trovärdighet. / Believability within fictional worlds is something many strive for, but how does one go about to achieve it? Movies and literature use the concept of verisimilitude to define a work of fiction’s appearance of being true or real and with this, also define how believable it is. Originating from Plato and Aristotle's term mimesis, which refers to the imitation of nature, the term verisimilitude developed into its current form to be used as a way of analysis to better understand how believability may be achieved. To do this, the term points to how well an audience can relate the fictional world to our own world, for if this is done well, it instantly becomes more believable. The term verisimilitude is mostly used within the works of movies and literature and as such, there is very little research to be found in relation to its use in video games. We argue that the incorporation of verisimilitude within video game design can prove to be a valuable tool to aid in how we design towards more believable worlds in video games. We will in this bachelor thesis inquire the help of Anneli Lethsalos studies on verisimilitude within movies and genres build an understanding of how verisimilitude is used within the films.We also inquire help from Kamal Bhatts studies on believability within computer games to research how believability in video-game differ from other mediums and what qualities a game needs in order to be seen as believable.  With this newfound understanding we then set out to design our own game environment with verisimilitude as a design perspective. With this we hope to establish a ground to how one can design around believability in video-games.

One way internet users determine the quality of a website is to look for so called 'credibility factors'. These factors can either be positive: the presence of a date, reference list, independent site certification; or negative: the presence of advertisements or broken links. This thesis investigates what role such factors play in the effectiveness of two e-health interventions. An e-health intervention is a health related website designed to change a person’s behaviour. Until now research into credibility has been largely theoretical. Studies have relied on subjective outcome measures such as Likert scales, website content recall, expressions of preference and self reported behaviour. This thesis describes two studies, the second of which investigates, for the first time, whether surface credibility manipulations change objective behavioural outcomes. Surface credibility is how much a perceiver believes a website on simple inspection. Based on a comprehensive literature review of credibility research, the following credibility factors were explored: presence of advertising, recognisable logos, contact details, physical address, references, third party certification, currency information, privacy statement, HTTPS encryption, top level domain and presence of a broken link. The first study involved the assembly of an exercise promotion website. Participants were randomised to receive the site modified to contain either factors heightening credibility or those lowering credibility. Participants using the high credibility version spent twice as long browsing the site as those using the low credibility version. There was no effect on attitude to exercise or self reported physical activity. The second study used the same methodology but with a website targeting an objectively measurable health behaviour (registration as an organ donor). In this study 889 university students were exposed to a website promoting organ donation. Information on the site was assembled based on theoretical domain interviewing and current research into organ donation interventions. 336 (37.79%) participants registered through the study website. The study detected no significant difference in registration rates between high and low credibility versions of the site. Of the 17 comments left on the low credibility site, only 3 were credibility related criticisms. It is the finding of this thesis that university students are willing to submit personal information and place trust in a website contravening many current credibility guidelines. Future studies into credibility are needed to explore why this is the case. One possibility is that the website was trusted simply because it was part of a research study. Another possibility is that the high quality of the textual content compensated for the lack of credibility of the site itself. It is the recommendation of this thesis that future studies focus on objective behavioural outcome measures and control for other forms of credibility such as participation in a research study.

In the tradition of acceptance and commitment therapy (ACT), cognitive fusion is a transdiagnostic risk factor and occurs when one becomes overly attached to or "caught up" in their thoughts, leading to a more narrowed behavioral repertoire and difficulty taking effective action in response to life's demands. Cognitive defusion is ACT's curative answer to fusion, and denotes the process of taking a step back, seeing thoughts as "simply thoughts," thereby reducing the negative impact of distressing or anxiety-provoking thoughts. While these components have been widely studied in neurotypical (NT) samples, the purpose of this study was to extend findings to people diagnosed on the autism spectrum (AS). Specifically, this study aimed to examine the impact of cognitive fusion in this population and the effectiveness of a brief defusion technique. Forty-two AS participants and fifty-five neurotypical participants were given a battery of questionnaires measuring psychological distress and dispositional levels of cognitive fusion. Participants were then randomized into either a brief cognitive fusion technique or a brief active distraction technique. In both conditions participants chose a distressing thought and rated it on a visual analogue scale (VAS) in terms of thought discomfort and believability. They were then read a rationale regarding their assigned technique, practiced the technique, and applied the technique to their chosen distressing thought. After the intervention participants immediately re-rated the thought on the same VAS. Throughout the study, participants' heart rate and skin conductance were monitored to determine physiological effects of the conditions. Finally, a follow-up survey was sent at a one-week and two-week follow-up, where participants re-rated the believability and discomfort of their thoughts. Results of this study showed that the AS group had higher overall levels of fusion than the NT group, and that fusion was moderately to strongly related to psychological distress in the AS group and the NT group. In terms of the intervention effects, all interpretation statements must be taken with caution, as there were significant pre-group differences despite randomization. Both defusion and distraction worked equally well in immediately reducing thought believability and thought discomfort for AS and NT groups. Furthermore, treatment effects were maintained at the two-week follow-up period for all groups except for the AS group in the defusion condition. There were no treatment effects for physiology. The current study provides evidence that cognitive fusion may be an important factor in the psychiatric comorbidity that people with AS experience, and a brief technique can be effectively used.

Commercial video game developers constantly strive to create intelligent humanoid characters that are controlled by computers. To ensure computer opponents are challenging to human players, these characters are often allowed to cheat. Although they appear skillful at playing video games, cheating characters may not behave in a human-like manner and can contribute to a lack of player enjoyment if caught. This work investigates the problem of predicting opponent positions in the video game Counter-Strike: Source without cheating. Prediction models are machine-learned from records of past matches and are informed only by game information available to a human player. Results show that the best models estimate opponent positions with similar or better accuracy than human experts. Moreover, the mistakes these models make are closer to human predictions than actual opponent locations perturbed by a corresponding amount of Gaussian noise.

No description available.