Search

The trait approach in the entrepreneurship literature which suggest that entrepreneurs possess risk-taking behavior as a personal attribute has yielded mixed and, in many cases, contradicting results. Subsequently, many scholars, through the cognitive approach, have postulated that entrepreneurs may perceive risks as challenges through a 'rosy lens', while others posit that entrepreneurs may overlook the risks inherent in business opportunities. This disregard of risk changes the perceptions of gains and losses upon opportunity evaluation, consequently making business opportunities seem more desirable than they really are. This notion suggests that entrepreneurs may very well be overly optimistic and misguided by a 'get rich or die trying' attitude that can be highly susceptible to systematic biases. The lack of empirical evidence in previous studies in this risk-perception-based approach represent a gap in the literature which this study seeks to address. Considering entrepreneurship as a process rather than an event, the present study focuses particularly on the pre-foundational phase of new venture creation. The conceptual framework developed seeks to examine the influence of risk perception on the formation of entrepreneurial intentions through its antecedents (i.e., attitudes toward behavior, subjective social norms, perceived behavioral control). Additionally, the baseline line model is then subjected to the moderating effect of cognitive biases (i.e., cognitive deviations from rational thought). By focusing on the influence of two biases, the illusion of control and overconfidence, the study's main theoretical assumption is that cognitive biases will strengthen the indirect relationship between risk perception and entrepreneurial intentions. The findings reached suggest that cognitive biases and risk perception play a strong role in forming entrepreneurial intentions. Indications show that risk perception and entrepreneurial intentions have a significant indirect relationship through attitudes and perceived behavioral control (PBC). In the collectivist culture of Saudi Arabia, no significance has been found for the effect of subjective social norms on potential entrepreneurs. The study contributes to the entrepreneurship literature by considering a multidisciplinary approach that is theoretically driven by intention models (e.g., the theory of planned behavior) and prospect theory.

The attention that academia and forensic agencies have been giving to the decision-making process of fingerprint examiners has increased in recent years. This is largely due to the increasing awareness that experts' decisions can suffer from cognitive bias, promoting erroneous decisions. Even though fingerprint examiners are expected to not commit errors deliberately, it is of interest to understand in detail the influences (i.e. impact on performance) that external factors such as contextual information have on their decision-making process. This thesis begins by investigating how a group of experts (n = 41) and another of laypeople (n = 57) differ in their accuracy and response time regarding the influences of different contextual information. Although experts showed higher levels of accuracy, findings suggest that both laypeople and experts have a tendency to suffer from similar types of cognitive bias associated with the same types of contextual information promoting lower accuracy and higher response times. It seems that different types of contextual information have different types of influence in experts' performance. Hence, this thesis analysed in a second study the accuracy and response times of 67 fingerprint examiners, from 15 forensic bureaus, based in 9 countries, during trials that simulated the Verification phase of the ACE-V process. Results showed significant differences between specific types of contextual information when compared with control trials, challenging guidelines that suggest fingerprint examiners should work within a full blind setting. Due to the diverse sample in this study, it was possible to explore differences regarding the different levels of experience, methodological approaches currently in use (i.e. numerical approach versus holistic approach) and the accreditation standards that forensic bureaus had. Findings provided insight that can be used in future developments for methodologies, accreditation guidelines, and training for fingerprint examiners. Finally, a qualitative study was conducted in which 42 fingerprint examiners were interviewed regarding their opinions about contextual information influences and the methodologies within the ACE-V process. Results retrieved from interviews shed light on aspects of experts' work, allowing a better understanding of the motivation and the level of cognitive enjoyment (assessed by the level of Need for Cognition) as well as the specific types of contextual information which may not influence experts' performance negatively, but rather be a motivational factor for their work.

Human activity is driving global biodiversity loss. However, the effects on animal cognition and emotion are less studied. In this thesis, I argue that anthropogenic change impacts animals' mental states, with implications for both individuals (welfare) and populations (conservation). My first experiment explores the transition from keeping dairy cattle at pasture to housing them indoors full-time. Using a repeated-measures crossover design, I gave cows three weeks of overnight pasture access and three weeks of indoor housing. Treatment did not influence judgements of ambiguous stimuli - a cognitive measure of emotional wellbeing. Nevertheless, behavioural welfare indicators (lying, walking, and anticipatory behaviour) suggested that subjects had more comfortable, rewarding lives at pasture. Next, I review attention bias, another potential cognitive indicator of animal wellbeing. Attention to threat proves a promising method to quantify the emotional impacts of anthropogenic stressors. I then investigate a second example of human-induced environmental change: oceanic microplastic pollution. Microplastic exposure prevented hermit crabs from approaching and entering a new shell, which was better than their current shell. These results suggest that microplastics disrupt animal cognition (resource assessment and evaluation). Finally, I apply emotion theory to animal contests, and argue that emotions underpin virtually all non-reflexive behaviour. Because emotions generalise across contexts, my novel approach suggests that human activity has broader psychological impacts than usually recognised. These findings highlight how anthropogenic change can influence animal cognition and emotion, with practical applications for welfare and conservation.

As new display technologies emerge and become part of everyday life, the understanding of the visual experience they provide becomes more relevant. The cognition of perception is the most vital component of visual experience; however, it is not the only cognition that contributes to the complex overall experience of the end-user. Expectations can create significant cognitive bias that may even override what the user genuinely perceives. Even if a visualization technology is somewhat novel, expectations can be fuelled by prior experiences gained from using similar displays and, more importantly, even a single word or an acronym may induce serious preconceptions, especially if such word suggests excellence in quality. In this interdisciplinary Ph.D. thesis, the effect of minimal, one-word labels on the Quality of Experience (QoE) is investigated in a series of subjective tests. In the studies carried out on an ultra-high-definition (UHD) display, UHD video contents were directly compared to their HD counterparts, with and without labels explicitly informing the test participants about the resolution of each stimulus. The experiments on High Dynamic Range (HDR) visualization addressed the effect of the word “premium” on the quality aspects of HDR video, and also how this may affect the perceived duration of stalling events. In order to support the findings, additional tests were carried out comparing the stalling detection thresholds of HDR video with conventional Low Dynamic Range (LDR) video. The third emerging technology addressed by this thesis is light field visualization. Due to its novel nature and the lack of comprehensive, exhaustive research on the QoE of light field displays and content parameters at the time of this thesis, instead of investigating the labeling effect, four phases of subjective studies were performed on light field QoE. The first phases started with fundamental research, and the experiments progressed towards the concept and evaluation of the dynamic adaptive streaming of light field video, introduced in the final phase.

Mental representations can be consciously experienced in mental imagery format, and verbal-linguistic format. Mental imagery representations of emotional information can evoke more powerful emotional responses than verbal-linguistic representations of the same information. Biases in mental imagery-based cognition are postulated to play a role in the maintenance of emotional disturbance in depression. Despite growing research, two questions remain: 1) is dysphoria (mild to moderate depression symptoms) associated with mood-congruent bias in the frequency of mental imagery generation; and 2) are such biases related to state emotional experience and emotional response to emotional information in dysphoria? To examine question one, participants varying in levels of dysphoria reported the occurrence of mental imagery in real time under task contexts that were emotional (negative and positive verbal cues) and unemotional (neutral verbal cues). Mental imagery generation was assessed under two task conditions: a) when participants were instructed to generate mental imagery in response to verbal cues (Study 1 & 2); and b) when participants were not instructed to generate mental imagery (or verbal-linguistic representations) during exposure to similar verbal cues (Study 2, 3, & 4). Results from all studies, across both instruction types, showed that dysphoria was associated with a loss of positive bias in mental representation generation, driven by reduced positive representation generation (Study 1, 2 & 4), but also by elevated negative representation generation (Study 1, 2 & 3). Interestingly, evidence of a loss of positive bias was most consistently observed when given neutral verbal cues, but also when given positive verbal cues. However, such dysphoria-linked effects were not disproportionately evident for mental imagery relative to verbal-linguistic representations, when both were allowed to naturally occur in Study 2, 3, & 4. Unexpectedly, dysphoria was associated with reduced tendency to generate negative imagery relative to negative verbal-linguistic representations in Study 2, though this finding was not replicated in Study 3 or Study 4. To examine question two, participants provided state mood ratings in addition to reporting mental representation occurrence during exposure to auditory emotional information (Study 3: verbal cues; Study 4: news stories). Dysphoria and mental representation generation was found to be unrelated to emotional response on negative trials (Study 3 & 4). However, greater occurrence of mental imagery, but not verbal-linguistic representation generation was related to greater positive emotional response on positive trials for individuals with dysphoria (Study 3), and all participants (Study 4). Study 5 analysed existing clinical trial data and found that the vividness of positive future event imagery is related to optimism in depression, such that those able to envision a brighter future are relatively more optimistic, and regain optimism more quickly, than those less able to do so, even when currently depressed. In summary, dysphoria was associated with loss of positive bias in mental representation generation, though such effects were not unique to imagery. Importantly, greater occurrence of mental imagery-based, but not purely verbal-linguistic, representations were associated with greater positive emotional response to positive information, and may hold value as a target for future translational research.

Animals are sentient beings, capable of experiencing emotions. Being able to assess emotional states in farm animals is crucial to improving their welfare. Although the function of emotion is not primarily for communication, the outward expression of an emotional state involves changes in posture, vocalisations, odours and facial expressions. These changes can be perceived and used as indicators of emotional state by other animals. Since emotions can be perceived between conspecifics, understanding how emotions are identified and how they can spread within a social group could have a major impact on improving the welfare of farmed species, which are mostly reared in groups. A recently developed method for the evaluation of emotions in animals is based on cognitive biases such as judgment biases, i.e. an individual in a negative emotional state will show pessimistic judgments while and individual in a positive emotional state will show optimistic judgments. The aims of this project were to (A) establish whether sheep and goats can discriminate between images of faces of familiar conspecifics taken in different positive and negative situations, (B) establish whether sheep and goats perceive the valence (positive of negative) of the emotion expressed by the animal on the image, (C) validate the use of images of faces in cognitive bias studies. The use of images of faces of conspecifics as emotional stimuli was first validated, using a discrimination task in a two-armed maze. A new methodology was then developed across a series of experiments to assess spontaneous reactions of animals exposed to video clips or to images of faces of familiar conspecifics. Detailed observations of ear postures were used as the main behavioural indicator. Individual characteristics (dominance status within the herd, dominance pairwise relationships and humananimal relationship) were also recorded during preliminary tests and included in the analyses. The impact of a low-mood state on the perception of emotions was assessed in sheep after subjecting half of the animals to unpredictable negative housing conditions and keeping the other half in good standard housing conditions. Sheep were then presented with videos of conspecifics filmed in situations of varying valence. Reactions to ambiguous stimuli were evaluated by presenting goats with images of morphed faces. Goats were also presented with images of faces of familiar conspecifics taken situations of varying emotional intensity. Sheep could discriminate images of faces of conspecifics taken either in a negative or in a neutral situation and their learning process of the discrimination task was affected by the type of emotion displayed. Sheep reacted differently depending on the valence of the video clips (P < 0.05); however, there was no difference between the control and the low-mood groups (P > 0.05). Goats also showed different behavioural reactions to images of faces photographed in different situations (P < 0.05), indicating that they perceived the images as different. Responses to morphed images were not necessarily intermediate to responses to negative and positive images and not gradual either, which poses a major problem to the potential use of facial images in cognitive bias experiments. Overall, animals were more attentive towards images or videos of conspecifics in negative situations, i.e., presumably, in a negative emotional state. This suggests that sheep and goats are able to perceive the valence of the emotional state. The identity of the individual on the photo also affected the animals’ spontaneous reaction to the images. Social relationships such as dominance, but also affinity between the tested and photographed individual seem to influence emotion perception.

Diagnosis of Alzheimer’s disease is encouraged as a first step towards people planning for their future with the condition. Despite the proposed benefits of diagnosis, it is also widely recognised that Alzheimer’s disease can expose people to stigma. Therefore, this thesis explores the relationship between stigma and future outlook, from the perspective of people affected by early and late-onset Alzheimer’s disease. In order to recognise the physicality of the condition and how psychological and social factors influence experiences, a biopsychosocial perspective is employed throughout. People with Alzheimer’s disease (n=15 people with late-onset, 7 people with early-onset) and their supporters (n=22) completed questionnaires about perceived stigma. This was followed by 14 interviews with a subsample of participants, which explored stigma and future outlook in more depth. Perceived stigma reporting across participants was low in the questionnaires; whereas interviews revealed higher levels of stigma with people discussing mixed, unpredictable reactions from a range of sources. Participants expressed awareness of the unpredictable nature of their futures with the condition. The subsequent lack of control was managed through focusing on ‘one day at a time’ and avoiding looking too far ahead. Across reflections on stigma and future outlook there was a deliberate focus on positive experiences for people affected by early and late-onset Alzheimer’s disease. The similar management of experiences across participants minimised possible age-based differences. These findings are supported by socioemotional selectivity theory, which suggests people are motivated to maintain positive emotional states when facing ‘time-limiting’ conditions irrespective of age. The research suggests people’s experiences of stigma and future outlook interact, with stigma-driven assumptions about the future affecting how people manage their daily lives. The avoidance of looking ahead suggests that policy which encourages future planning should consider its utility and explore ways of helping people to manage both exposure to stigma, and planning for the future, whilst focusing on daily living.

The idea that humans are prone to widespread and systematic biases has dominated the psychological study of thinking and decision-making. The conclusion that has often been drawn is that people are irrational. In recent decades, however, a number of psychologists have started to call into question key claims and findings in research on human biases. In particular, a body of research has come together under the heading of adaptive rationality (henceforth AR). AR theorists argue that people should not be assessed against formal principles of rationality but rather against the goals they entertain. Moreover, AR theorists maintain that the conclusion that people are irrational is unsupported: people are often remarkably successful once assessed against their goals and given the cognitive and external constraints imposed by the environment. The growth of literature around AR is what motivates the present investigation, and assessing the plausibility of the AR challenge to research on human biases is the goal of this thesis. My enquiry analyses several aspects of this suggested turn in the empirical study of rationality and provides one of the first philosophically-informed appraisals of the prospects of AR. First and foremost, my thesis seeks to provide a qualified defence of the AR project. On the one hand, I agree with AR theorists that there is room for a conceptual revolution in the study of thinking and decision-making: while it is commonly argued that behaviour and cognition should be assessed against formal principles of rationality, I stress the importance of assessing behaviour against the goals that people entertain. However, I also contend that AR theorists have hitherto failed to provide compelling evidence in support of their most ambitious and optimistic theses about people’s rationality. In particular, I present a great deal of evidence suggesting that people are often unsuccessful at achieving prudential and epistemic goals and I argue that AR theorists have not made clear how, in light of this evidence, optimistic claims about human rationality could be defended.

The decisions made by Mental Health Professionals (MHPs) are of utmost significance for providing the highest quality care to service users. The assessment of risk is one of the pivotal processes that MHPs undertake frequently, as per government policy guidelines, and in order to safeguard patients and the public. Although Risk Assessment Proformas (RAPs) consume a proportion of MHP time and resources, very little research has been undertaken to address factors that might affect their most optimal utilisation in practice. Previous literature suggests that medical decisions, like decision making of other kinds, is fraught with difficulty including being susceptible to the influence of cognitive biases, pre-decisional affect, overconfidence, and subjectively held attitudes towards organisational policies and regulations. Specifically, the presentation of risk information can influence decisions. It has also been suggested that anxiety has the capacity to elicit risk aversive responses, and that overconfidence and negative attitude may lead to complacency in undertaking policy-led responsibilities and produce non-compliance for the same. However, much of what is known about medical decision making has been gleaned from outside of context of mental health. As such, the current programme of research aimed to explore decision making in mental health settings and with a view to raise awareness of the complexity of decision making amongst MHPs. The implementation of quantitative and qualitative techniques (studies 1 and 2) revealed negative attitudes from psychiatrists towards Risk Assessment Proformas (RAPs), which are essentially structured decision making aids. Psychiatrist, compared to other MHPs, spent less time completing RAPs, which may reflect their differing attitudes towards their usefulness, something that was consistently emphasised during in-depth qualitative exploration. It was also found that experience was an additional differentiating factor between MHPs. Relationships between experience and other factors such as anxiety, confidence and complacency were found via conversations with MPHs, experience members of staff being less inclined to provide comprehensive and detailed accounts of service user risk in RAPs. This is problematic since although there is, in the UK, a policy led requirement that RAPs are completed for each service user, it is clear that there are inter-professional variations in how RAPs are being used and this acts to inhibit the best information sharing between all those involved in patient care. Following previous work in the area of cognitive bias and its influence upon general and medical decisions, a clinical vignette was also developed (study 3) to establish whether the presentation of risk information influences psychiatric admission decisions. The current findings supported previous work in that decisions were susceptible to the framing effect. The findings here, and previously in the literature, reveal a necessity for MHPs to be informed of bias in decision making in an attempt to improve objectivity in risk assessment practices. The unearthing of the framing effect also further signals the need for proper use of RAPs, where many MHPs may not be using them to their full potential - i.e. an aid to the systematic consideration of a range of information about a service user. The final part of the thesis (study 4) turned to the piloting of an educational module incorporating content around the factors affecting decision making in an attempt to raise awareness amongst MHPs. The rationale being better awareness of the complexity of decision making may act to enhance decision making processes. Pre and post intervention analyses revealed an improvement of baseline to follow-up knowledge of decision making bias and statistical concepts and this knowledge was maintained to a moderate level at four weeks follow-up. Although individuals maintained their susceptibility to the framing effect, the bias was less prevalent in those who knew of its presence before taking part in the study. Overall the findings give some support to the use of education as an approach to raising awareness about decision making processes in MHPs, although what remains to be seen is whether such education acts to bring about changes in behaviour - for example, different use of RAPs. The PhD programme suggests that MHPs are just as susceptible to cognitive biases, such as the framing effect, as has been demonstrated in both general population and other groups of health practitioners. At the same time, attitudes to RAPs differ depending on exact job role, which psychiatrist being least likely to spend time on their completion and reporting them as a tool for noting decisions reached as opposed to an aid to the process. This acts reduce the quality and quantity of reported information shared with colleagues about a service user. It is possible that MHP behaviour aligns with general attitude-behaviour models, such as the Theory of Planned Behaviour. As such, whilst the current work has demonstrated that educational interventions may act to improve awareness of decision making processes and their influences, further research would benefit from considering if these types of approach affect actual behaviour. For example, improved used of RAPs as decision-aids, reduced susceptibility to framing effects, consciousness around how information is represented in RAPs given knowledge of how the information may be used by others.

Varying internal and external stimulants naturally oscillate the creation of objective intelligence analysis; however, the most egregious offender to accuracy and completeness is the natural human mental process. No element of intelligence analysis is more formidable than the cognitive process that houses specific deviations, known as biases, as it yields inaccuracies and alters what is believed to be a rational response to a complex analytical problem. The phenomenon of ethnocentrism is consistently identified as an analytical limitation of intelligence professionals, derived from cognitive bias. Ultimately, ethnocentrism manifests an analyst’s perception of information directly through the lens of culturally dependent heuristics and cognitive patterns accumulated over a lifetime. The application of the Six-Dimensions of National Culture (6-D model), put forth in 1980 by Geert Hofstede, was selected to alter this pattern via an exploratory qualitative multi-case study involving Imperial Japan (Pearl Harbor), the Soviet Union (Cuban Missile Crisis), and al-Qaeda before September 11, 2001. With the application of the 6-D model, in concert with the Data-Frame Theory, as presented by Moore and Hoffman, this dissertation attempts to amplify the human capacity to make sense of ethnocentrism by expanding analytical frames and, as a result, help analysts produce and disseminate more holistic intelligence. Research results identified significant divergence between the created cultural disposition and the US intelligence perspective, specifically throughout the IDV, MAS, and IVR cultural dimensions. This study highlights the implications of ethnocentrism within the intelligence paradigm and identifies the necessity to employ cultural condition frameworks when attempting to produce accurate and comprehensive intelligence.

Welfare for captive animals in laboratory, farm, companion and zoo settings is difficult to assess and subject to interpretation. The perceived needs of animals change over time and according to budget, fashion and policy. The assessment of the value of environmental enrichment for captive animals is hampered by the lack of consistent, quantitative, objective and methodical research. Enrichment devices and other welfare improvement strategies lack long-term assessment or implementation. The concept of cognitive enrichment, or enabling captive animals to use their cognitive abilities to solve problems and meet challenges in their environment, is in the initial stages of development. Preliminary findings indicate that cognitive enrichment seems to be an effective method of improving captive animal welfare. In the research described in this thesis, several welfare measurement techniques including changes in activity budget, stereotypic behaviour and cognitive bias were compared and used to measure the effects of cognitive enrichment upon captive pigeons and dogs in order to obtain a useable cognitive enrichment paradigm that can be extrapolated to many types of captive animals. Key findings were that cognitive enrichment improved captive group-housed pigeon welfare and individually-kennelled dog welfare, indicated by significant changes in activity budget, increases in ‘optimism’, and reduction of stereotypic behaviour in subjects. Subjects used both low-tech and high-tech cognitive enrichments as intended and showed no sign of habituation. It is hoped that these findings will be used to improve captive animal welfare, and that the cognitive enrichment and cognitive bias paradigms developed will add to this field of research.

Child and adolescent mood and anxiety symptoms are common and debilitating, with long-term effects on well-being. Research presented in this thesis examines interpersonal cognitive factors in the emergence of anxious and depressive symptoms in late childhood through to early adulthood. The thesis considers this issue using three main approaches. For the first, data are presented showing that biases in the appraisals of social situations are the aspects of interpersonal cognition most closely associated with emotional symptoms. For the second, longitudinal twin data are used to examine genetic and environmental origins of these interpersonal cognitive biases and their temporal prediction of symptoms across a 2-year period. Data show that interpersonal cognitive factors are strongly influenced by non-shared environmental factors, and moreover, predict symptoms across time. The final section of the thesis comprises four studies using Cognitive Bias Modification of Interpretations (CBM-I) training methodology to show that both positive and negative interpretive biases for interpersonal information can be induced in adolescents. Positive biases are shown to persist for at least 24 hours after training, and induced positive and negative biases are shown to differentially predict anxious responses to an experimental stressor. Evidence is also provided to suggest that effects following training positive interpretive biases may transfer to other cognitive measures, namely appraisals of ambiguous emotional faces. Finally, data tentatively show that CBM-I training may be useful in reducing negative interpretations of interpersonal information made by 11-year-old children undergoing the transition to secondary school. In summary, studies in this thesis support the contribution of cognitive biases to mood and anxiety symptoms in childhood and adolescence. They further extend this knowledge by suggesting that these reflect individual-specific (non-shared) environmental risks to predict symptoms across time. These biases may also be amenable to change through training interventions, with some - albeit weak - effects on other cognitive outcomes.

Patients with both depression and anxiety show an increased tendency to deploy attention towards negative information. Cognitive models of the illnesses predict that these negative attentional biases are causally related to the symptoms of the disorders. Consistent with this, modifying attentional bias using either antidepressant medication or simple, computer based training tasks has previously been associated with altered symptomatology in both non-clinical and clinical populations. The current thesis aimed to investigate the mechanisms by which attentional bias training tasks alter attention. The investigations were conducted within an experimental neuroscience framework which has previously been successfully deployed in studies of antidepressant medication. The thesis then sought to use these initial results to improve the basic understanding of attentional control processes and, ultimately, guide the development of novel treatment strategies. The initial studies of the thesis characterised the behavioural and neural effects of attentional bias training. Behaviourally, a high degree of generalisation of the training effect was found across a range of emotional stimuli. Neurally, training was found to alter activity in a network of prefrontal regions known to be involved in the control of attention. Further analysis, utilising a computational learning model, suggested that the attentional control systems identified in this study could be understood in terms of expectation based processes. These studies therefore indicated that, in contrast to the predominately limbic effects of antidepressant medication, training initially altered the response of frontal control circuitry. The later studies of the thesis investigated possible strategies for extending the use of attentional bias training. Firstly, combining training with antidepressant medication was found to produce an interference effect on emotional memory suggesting that administering both interventions concurrently is likely to erode their cognitive impact. Lastly, attentional bias training was found not to alter attention in patients with bipolar disorder, with the results of the study indicating that standard assessments of attentional bias in this clinical population are likely to be unreliable. Overall, these studies indicate that attentional bias training may be used to alter the top-down control of attention to emotional information and suggest that such effects may interfere with the bottom-up effects of antidepressant drugs. More generally the work demonstrates the utility of using a cognitive-neuroscientific framework to explore the mechanisms and impact of novel therapeutic strategies.

This thesis presents the development and application of methods to assess cognitive markers of emotion and psychological wellbeing in a species of nonhuman primate, the rhesus macaque (Macaca mulatta). In humans, vulnerability to emotional disorders such as anxiety and depression is characterized by particular cognitive profiles, known as cognitive biases. For example, anxious people automatically attend to threat-relevant information, interpret ambiguous information negatively, and have negative expectations of future events. In this thesis, I first describe two treatments that were used prior to cognitive testing to induce positive and negative shifts in inferred affective state in the monkeys (enrichment and a health-check, respectively) and discuss the impact of these treatments on the monkeys’ behaviour and physiology (Chapters 2 and 3). In the first cognitive study (Chapter 4), I present a method that uses eye-gaze to assess the extent to which threatening (versus non-threatening) stimuli capture visual spatial attention when two stimuli are presented at different locations. In the second study (Chapter 5), I present a simple operant touch-screen task to assess the extent to which a threatening distractor stimulus captures attention and impairs performance on an ongoing task when presented at the same location as the taskrelevant stimulus. In the third study (Chapter 6), I present a Go/NoGo touchscreen task to assess judgements about the reward value of ambiguous stimuli. In all of these studies, the two treatments led to different cognitive profiles in the monkeys. Monkeys showed a) automatic capture of attention by threatening stimuli, which was followed by avoidance following the health-check, but not Post-enrichment; b) impaired task performance when a threatening distractor stimulus was presented Post-health-check, and improved performance on these trials Post-enrichment; and c) a more negative judgement about the reward value of ambiguous stimuli Post-health-check versus Post-enrichment. I discuss these cognitive biases in light of available data from humans, and recent work with nonhuman animals. These data indicate that furthering our understanding of primate and other animal psychological wellbeing, may be achieved through the development of measures of cognitive bias, such as those presented here.

A status quo bias is a tendency to resist change and keep things as they are. This bias is robust in humans and likely a byproduct of heuristic decision-making mechanisms, indicating that it may be an evolutionarily-conserved process that is phylogenetically widespread. However, few status quo bias studies have been conducted with nonhuman animals, and the evidence was mixed. Studying this question with animals could help inform welfare decisions for animals directly and it can shed light on the degree to which a status quo bias (as seen in humans) may be a result of human-unique experiences or the consequence of more fundamental decision-making mechanisms. The goal of this study was to explore whether other primates exhibit a preference for the status quo after controlling for reinforcement history, and to explore the relative influence of several different factors that may moderate the status quo bias effect. To test these questions, I conducted two experiments. In Experiment 1, capuchin monkeys and macaques made choices among computerized tasks, where one task was presented as the ‘default’ option (i.e., it could be played continuously without having to select it from a menu), after forced runs of trials that varied in length and which included trials of either a single task type or a mix of task types. I predicted that the monkeys would demonstrate a general preference for the default task over non- default alternatives, and that this preference would increase in magnitude after longer forced runs of the default task. In Experiment 2, lemurs and tamarins learned that boxes could be opened in two ways (lifting or sliding). I explored whether animals could be influenced to open the box by using one particular mechanism after recent exposure to that mechanism (i.e., establishing it as the ‘status quo’) while controlling overall reinforcement history for both mechanisms. The results indicated that the animals did not exhibit a status quo bias in these paradigms; other factors such as variety and task preference (E1) and approach angle (E2) had greater influence on animals’ choice behavior than the established status quo.

For many early-stage investors, startup valuations are more than science. Due to the information asymmetry, uncertainty, and risk involved, valuations are complex to understand and perform. This mixed-method study explores how investors value new tech ventures when they cannot use analytical methods. We open-ended interviewed fifty-eight investors and analyzed nine hundred seed investment rounds between 2010 and 2020 around the globe. Using qualitative and quantitative techniques, we found that investors are limited in using valuation methods, leaving the valuation process to investor intuition, mental models, prior assumptions, and judgment, causing mental shortcuts and cognitive biases to distort the process. The scarcity effect, signaling, herd mentality, and impact of stereotypes such as entrepreneur ́s school attended, race, gender, and region were found as variables that affect the startup valuation and money raised. Such results are supported by the behavioral economics literature and bring attention to a different perspective to understand a startup's valuation by analyzing the investor behavior.

Smart technology, such as the internet of things, artificial intelligence, and big data, provides consumers with a new level of convenience through various smart-connected products (SCPs). Although many experts have increasingly warned about the privacy vulnerability issues of various SCPs, consumers often underestimate privacy risks when adopting smart technology. Accordingly, this dissertation presents a literature review and three empirical studies that examine the privacy problems and suggest new concepts and models for a deeper understanding of the privacy suspension phenomenon. The first chapter reviews the literature related to the privacy suspension phenomenon by integrating the antecedents of consumers' privacy concerns. New concepts of privacy concerns, such as active and inactive privacy concerns, are suggested along with multiple propositions for the proposed privacy suspension theory, which extends the dimension of ambivalence toward trust and distrust regarding smart technology. The second chapter presents the proposed privacy–common good trade-off model and three assumptions related to privacy trade- offs, privacy reduction, and anchoring effects in the sustainable smart-connected car context. This study also discusses the relationships between governments, companies, and consumers regarding the effects of the common good of sustainability and government subsidies. The third chapter evaluates the mediation effects between sustainability, trust, privacy concerns, disclosure intentions, and purchase intentions when purchasing sustainable smart-connected cars based on the proposed sustainability–trust–behavior model. Finally, the fourth chapter provides a practical solution to resolve privacy suspension issues using the design science research approach. This study proposes privacy information type characteristics to evaluate SCPs' tailored data collection capabilities, visualizing them through a spider diagram design method with nudges.

Bears are charismatic, long-lived animals held by nearly half of AZA zoos, and yet pressing questions remain about how the zoo environment affects their welfare. The goal of this thesis was to investigate this from a variety of perspectives to reveal specific factors that may affect bear welfare. Pacing is a commonly used welfare indicator which lacks a standard operational definition. After reviewing the literature and conducting a comparative analysis of commonly used methods, I recommended instantaneous sampling and the ABA definition as a standard method for ease of use, consistency, and ability to compare to previous studies. Seasonal environmental factors such as changes in temperature, light, and weather, as well as the presence of zoo visitors, and management strategies such as access off-exhibit and habitat rotation may affect bear behavior. I established a longitudinal bear monitoring program at CMZ, and these investigations revealed that seasonal changes, daily attendance, crowd size at the habitat, and a temporary Zoo closure had limited effects on bear behavior, and that carefully monitored management strategies can improve both bear welfare and visibility. As there are relatively few validated positive welfare indicators for bears, I investigated the use of cognitive bias as a novel measure of positive welfare. Two grizzly bears were trained to respond to a judgment task using a touchscreen computer during four experimental conditions. This task revealed an underlying optimistic affect in one bear in one condition, suggesting that this type of testing is a promising direction for future studies. Finally, to gain a holistic view of how one AZA zoo’s approach to welfare assessment has functioned, I conducted a preliminary evaluation of the Cleveland Metroparks Zoo’s animal welfare process, as it has been applied to bears. I used a mixed-methods approach to investigate whether this process met the goals of improving documentation, communication, and responses to welfare concerns. This evaluation revealed that overall, the process has contributed to optimizing bear welfare by improving communication and management responses to welfare issues. These studies provide a holistic examination of bear welfare in the zoo setting.

Supplementing one’s pre-existing knowledge by accessing information (e.g., answers to tests) is quite common and subsequently inflates a participant’s estimate of their internal knowledge (Fisher et al., 2015). The purpose of this research is to examine whether access to answers promotes the Better-Than-Average Effect, or BTAE (Alicke, 1985). The BTAE describes a person’s tendency to rate themselves as above average on various highly desirable traits and characteristics; this is one form of self-enhancement by which people maintain positive views of the self (Leary, 2007). While access to information has been found to inflate a participant’s estimate of their internal knowledge, the present research extends these findings by examining whether such answer-access impacts how participants perceive their level of knowledge – and performance – as it compares to others. Experiment 1 hypothesized that compared to participants without access to answers (i.e., from Google), participants using the Internet to complete a trivia task would rate themselves as more knowledgeable than either an individual ostensible participant or an “average college student” and, in turn, cause an overestimation of performance without Internet use. A follow-up experiment (Experiment 2) explored whether access to answers during completion of a vocabulary quiz inflates estimates of internal knowledge and in turn stimulates an overestimation of one’s performance on a hypothetical subsequent task; it also looked to extend Better-Than-Average Effect research (Alicke, 1985) by exploring whether access to answers inflates social comparison ratings on nine highly desirable traits and characteristics that were significant or marginally significant in the first experiment.

Cognitive bias refers to the influence of affective state on the interpretation of ambiguous stimuli and has been used to assess emotional state in nonhuman animals. The current study assessed cognitive bias in 12 brown-tufted capuchin monkeys using three distinct computerized psychophysical tasks and a novel manipulation to affect that involved giving moneys gelatin foods that tasted either pleasant or unpleasant. In addition, monkeys were trained on several positive and negative training cues. Results showed that food type was not a factor in monkeys’ responses to ambiguous stimuli. Behavioral observation during test sessions revealed the unpleasant food may have acted as a form of enrichment, thereby providing the monkeys with two pleasant activities prior to assessments of their emotional states. Further, results indicated that monkeys displayed a preference for the positive response class when classifying the ambiguous probe, but that this preference was subject to both task and individual differences.

Over 70% of adults will, at some time in their lives, lack decisional capacity and needanother person to make health-related decisions on their behalf (Silviera, Kim & Langa,2010). As advancing age increases the increasing risk of incapacity due to a suddenneurological event (e.g., stroke) or a progressive dementia (e.g., Alzheimer’s Disease),one would further presume that older adults would be even more motivated towarddiscussion of preferences for future healthcare and medical treatment, designation of aperson to serve as a future surrogate health care decision-maker, and documentation ofpreferences for future healthcare and medical treatment). But, one-quarter of all adultswill not have completed any of these advance care planning behaviors prior to death(Bischoff, Sudore, Miou, Boscardin & Smith, 2013). This study applied the MotivationalModel of Aging and Decision-Making Competence (Strough, Parker & Bruine de Bruin,2015) to examine the potential influence of stable attributes of older adults, deliberativemotivation, deliberative cognition, and cognitive bias in advance care planning for futuredecisional incapacity among a sample of 266 older adult participants of the 2012 cohortof the Health and Retirement Study. Overall, using the selected measures, the model didnot predict advance care planning behavior in this sample. Analyses of the individualcomponents of the process did indicate a number of meaningful relationships that warrantfuture research.

The rapidly increasing size of data is becoming a major challenge for both humans and machines to process. While more data means more information and less uncertainty, and consequently better performance, more data also means more processing time and more storage. This motivated my thesis which is centered around finding ways to cut the size of data shown to humans, as well as data fed into machine learning algorithms, without compromising the performance. First, in the context of human decision making, we aim at reducing and reordering the data to show to a human subject to enhance their decision performance. We propose a statistical model for human decision making that incorporates cognitive biases. We then propose an algorithm that allows, in polynomial time, to construct an ordered subset of the data so that the human performance approximately matches the optimal performance. Second, we propose an algorithm for selecting a subset of the training data to train the SVM on. The algorithm optimizes a submodular set function, that represents the diversity and the relevance of the subset considered, while providing some performance guarantees. We then propose an algorithm for selecting a weighted subset of the training data to train the SVM on. The weighted subset construction is based on constructing the maximal independent set of the graph induced by the approximate nearest neighborhood properties of the dataset. Third, we propose two algorithms for online selective training for neural networks. The first method is based on picking batches that maximize the reduction in entropy of the estimator. The second method consists of constructing the batches such that all the datapoints included have predicted probabilities under some threshold. Our approaches allow to keep the epoch based framework of training neural networks, and to make the decisions based on up to date values.

There is increasing recognition within the forensic science community that human examiners are prone to biases that may affect the accuracy and precision of conclusions. The analysis of decomposition to estimate the postmortem interval (PMI) is one of the important functions of forensic anthropologists, yet analytical methods that rely on the judgment of human observers, such as assessing PMI from decomposition, may be vulnerable to cognitive bias, leading to inaccurate results. No studies to date have examined the effects of cognitive bias on decomposition scoring methods. The goal of this dissertation is to understand the role of the cognitive factors of mood and motivation on the analysis of decomposition characteristics both in field and photograph contexts. Fifty undergraduate research assistants with the University of Tennessee’s Forensic Anthropology Center assessed decomposition from 10 donated individuals over several months utilizing the Total Body Score method. Four graduate students who had extensive knowledge and experience with decomposition served as experts with which to compare the accuracy of observers’ TBS values. Observers also completed two psychological measures aimed at assessing their motivations (Intrinsic Motivation Inventory) and mood (Positive and Negative Affect Schedule) at the time of scoring. It was expected that observers and experts would differ in both the field and photograph contexts and that mood and motivation factors would be predictive of TBS values. Hierarchical random intercept multiple regression models were conducted to assess the relationship between these measures of cognition and observers’ TBS scores. Observers and experts only differed in the field context possibly due to biasing contextual information regarding the placement date of the donors. No differences were found in the photograph contexts because contextual information was not available to observers. Observers in the field context were motivated by their perceived performance, how nervous or pressured they felt while completing the task, and the perceived difficulty of the task itself. Mood only played a small role in the photograph context where negative mood influenced scoring decisions. This study has implications for the application of TBS to longitudinal research conducted at outdoor decomposition facilities as well as to cross-sectional casework.

Rationale. Treatment for children with obsessive compulsive disorder (OCD) involves deliberative approach of feared situations via exposure and response prevention, however a substantial proportion of youth do not respond. One potential explanation for the failure of deliberative approach of feared situations to result in complete symptom reduction may be that patients also have prepotent, or automatic response tendencies to not approach or to avoid feared stimuli. Measures of automatic approach and avoidance biases have been developed and examined using standardized pictorial stimuli among populations with fairly homogenous fears. However, the highly heterogeneous nature of OCD symptoms suggests the need to utilize idiographic stimuli. The current pilot study represented an attempt to demonstrate initial feasibility, reliability, and preliminary validity of an idiographic measure of automatic approach and avoidance biases among children diagnosed with OCD. Design. Participants (N = 17) included children ages 8-16 who were enrolled in two larger intervention studies. Symptom severity was assessed via clinician, child, and parent ratings from the Children’s Yale Brown Obsessive Compulsive Scale. With the assistance of a clinician, parents and children worked collaboratively to identify threat pictures that triggered the child’s OCD obsessions and/or compulsions, as well as emotionally neutral pictures. These pictures were utilized in an Approach Avoidance Task (AAT). In this computerized task, children were presented with threat and neutral pictures on the screen one at a time, framed by a blue or green border. Participants were instructed to respond by pushing or pulling a joystick based on the color of the picture border, which resulted in the picture becoming progressively smaller (simulating avoidance) or larger (simulating approach), respectively. Reaction times were measured for each combination of stimulus type and response direction, thus allowing for calculation of bias scores for different stimulus-response combinations. Aim 1: To characterize the nature of approach-avoidance tendencies in children with OCD. Aim 2: To examine the relationship between approach-avoidance tendencies and OCD symptom severity.Results. All participants were able to identify an appropriate number of threat and neutral pictures for use in the task. Split-half reliability coefficients for individual reaction time measures were high, as were correlations between individual reaction time measures. When examining bias scores, reliability was acceptable only for approach bias (pull threat minus pull neutral). Participants were significantly faster to pull threat pictures toward themselves when compared to reaction times for pulling neutral pictures toward themselves (approach bias). Moreover, participants demonstrated a larger approach-avoidance bias (i.e., reaction time difference for pushing minus pulling pictures of a given stimulus type) for threat compared to neutral stimuli. Both of these effects were statistically significant in the opposite direction as hypothesized. Participants did not demonstrate a statistically significant avoidance bias (push threat minus push neutral). No correlations between bias scores and OCD symptom severity reached statistical significance. Conclusions. The current study demonstrates the feasibility of collecting idiographic stimuli from youth diagnosed with OCD, which had not yet been established in previous studies. Despite high reliability of individual reaction times, reliability of the bias scores may have been reduced by high correlations between individual reaction time means. Support for the hypothesized bias effects was limited. Establishing reliable measures of automatic approach and avoidance biases represents a critical area for continued research and a necessary step before firmer conclusions can be established about the nature of such biases or their relationship to symptoms.

The endowment and the framing effect are widely examined cognitive biases. The experimental economics literature, using choice data gathered through an elicitation device, commonly finds evidence of these biases. Recent work by Cason & Plott (2014) shows that the interpretation of choice data as consistent with biases related non-standard preference theory could also be consistent with confusion or misconception of the game type used to elucidate preferences. I use the Cason and Plott card auction framework to analyze offers of buyers and sellers in an experimental setting with subjects from the University of New Orleans simulating 97 sellers and 90 buyers. The two games have symmetric payoffs in order to examine cognitive biases given subjects’ misconception of the game form. Subjects of both games display misconception of game form that explains both endowment and framing effects by rational confused choice; however, buyers display a much greater dispersion of offers than sellers. I estimate card implied valuation of sellers and buyers given game form misconception and find no statistical endowment effect, but I do find valuation is more uncertain in the buyer’s game. The theory of Rational Inattention predicts this lack of offer symmetry is due to the additional cognitive steps necessary in calculating buyer offers.

In the face of uncertainty, human judgment and decision-making often tends to deviate from the realm of rationality. Gambler’s fallacy and its opposite, hot outcome, are two such departures from laws of probability involving random streak of events. However, the adaptive thinking approach to decision-making proposes that any irrational heuristic or illogical belief can be fully adaptive as long as it fulfills the requirements of the decision task. During penalty shootouts in association football, goalkeepers face a series of multiple penalty kicks which are independent draws from a random process, and they need to anticipate the likely kick directions with limited time, insufficient information and computational capacity. The objective of this current study was to observe the goalkeepers in real-world competitive settings and examine whether they use gambler’s fallacy and hot outcome as predictive strategies to decide on their dive directions during penalty shootouts following streaks of correct and incorrect predictions in the same direction. Another goal was to investigate from the adaptive thinking perspective whether such strategies lead to more correct predictions by the goalkeepers. Penalty shootout data were collected from the elite soccer tournaments over the course of last 25 years (1992 - 2016) and after applying appropriate exclusion criterions, 405 penalty kicks were considered for the final analysis. Binomial tests revealed that following progressively longer streaks of correct predictions in the same direction, goalkeepers became increasingly more likely to dive in the opposite direction for the subsequent kick than would be expected by chance, a behavior consistent with gambler’s fallacy. However, neither gambler’s fallacy nor hot outcome type patterns were observed when analyzing dives of goalkeepers following streaks of incorrect predictions in a particular direction. On the other hand, results of the Fisher’s exact tests confirmed that both the fallacies failed to produce significantly more correct predictions from the goalkeepers and hence, they are mere fallacies. Goalkeepers’ belief in gambler’s fallacy highlights the biases associated with their real-world decision-making as this study does not support the adaptive use of such deceptive beliefs as predictive decision strategies.

In response to the historic Paris Agreement on climate change and to the Environmental Protection Agency’s recently finalized Clean Power Plan, economists and other climate policy experts have renewed the call for the United States to adopt a carbon tax. Opposition among the public presents a major obstacle. While a majority of the public supports government action on climate change, most people favor the use of “green” subsidies and command-and-control regulations—a fact that frustrates economists of all political stripes who contend that a carbon tax would be much cheaper and more effective. This Article argues that a cognitive bias known as opportunity cost neglect pervades the public’s thinking about climate policy instruments, causing people to ignore the hidden costs of subsidies and command-and-control and, for that reason, to support less efficient alternatives to the carbon tax. The Article will help proponents of the carbon tax better tailor their advocacy efforts. The Article also contributes to the burgeoning literature on behavioral public choice, which shows how the cognitive biases of political actors (including voters) influence the law. In addition, the Article points to the possibility of a Pyrrhic victory for conservative policymakers who oppose the carbon tax. Rather than averting major government action on global warming, defeating the carbon tax will very likely facilitate adoption of more costly substitutes that the public strongly favors as a result of cognitive bias. In that respect, the Article lends support to recent proposals by a small, but growing group of conservative scholars, who argue for a policy swap in which conservatives agree to a revenue-neutral carbon tax in exchange for support from environmentalists for abandoning the government’s current regulatory approach. The Article also suggests that conservative policymakers rethink their position on the carbon tax given that the states are currently considering which policy options to pursue in satisfying their respective obligations to reduce carbon emissions under the Clean Power Plan.

Government agencies have attempted to reduce energy consumption using executive orders, mandates, and agency policies. Despite these efforts, overall energy consumption of government facilities has not experienced significant energy reductions. Why haven't these efforts succeeded? The premise is that energy consumption decisions and their unintended outcomes contribute to this problem, and in this manuscript research focuses on cognitive bias, choice architecture, and decision making in relation to energy decisions answer this question. Potential impacts cognitive bias has on the decision maker is examined, and if it is possible to design better decision environments to account for cognitive bias and help decision makers maximize benefits (utility). This manuscript first examines the literature of cognitive bias, choice architecture, and government energy management, especially how these topics relate to meeting the country's energy goals. The next chapter examines cognitive bias that government facility managers encounter using qualitative analysis. In this study, the research indicates facility managers encounter loss aversion, risk aversion, choice overload, and the status quo bias during energy decisions. The last chapter examines applications of choice architecture, specifically attribute framing, to emphasize the utility maximizing choice of long term energy reductions over initial cost. This study found that decision makers did not see the utility of the energy efficient option without an intervention to draw their attention to the long term savings. Once the decision makers became aware of the potential savings, they chose the most efficient (and utility maximizing) option.

The present research examined whether the innominate acted as contextual information which resulted in a skewed scoring of skull morphological sex traits and pairs overall (both innominate and skull). Survey participants first assessed the sex of an innominate, then assessed the sex of a skull, using methods standard in the field. The sex of ten skulls and associated innominates, which served as contextual elements to introduce bias, were assessed in two surveys by 22 participants total. Male and female innominates and skulls were mixed and matched to test bias, though it was implied to participants that the paired elements were from the same individual. No significant bias was seen on the level of the skull. However, significant shifts in scores were seen on the level of the overall pair in three out of the five studied, indicating that the sex of the innominate was privileged over that of the skull. This is considered standard procedure in the field as the innominate is more reliable for sex estimation; thus, the findings of the present study are inconsistent with cognitive bias. However, the present study raises questions about the utility of the skull in the estimation of sex in human skeletal remains. While the innominate may be more reliable in sex estimation when both elements are present, this may lead to inattention to valuable information presented by the skull in particular contexts such as commingled burials.

Despite the education efforts of health organizations, federal regulators, and food producers on the benefits and safety of food irradiation, consumers demonstrate considerable misinformation and express resistance to purchasing irradiated food or accepting irradiation as safe food technology, even though irradiation can substantially reduce the incidences of foodborne illnesses that hospitalize or kill thousands of American each year. Consumers’ resistance to food irradiation has been shown to be related to safety concerns (He et al., 2005), resistance to new food technologies in general (Zachman & Østby , 2011), and balancing risks against benefits regarding contracting bacterial illness and irradiation (Eustice & Bruhn, 2007). The objective of this study is to examine how food labeling may mitigate cognitive biases about food irradiation, leading to more accurate beliefs about food irradiation treatment and ultimately to more positive attitudes and intentions regarding irradiated food purchases. This research shows that any labeling regarding irradiation places a stigma on the product. Labels that include bias-mitigation messages have a moderate effect on consumers’ acceptance of irradiated food.

When faced with a decision regarding probability or heuristics, people generally show their bias toward a heuristic, even if it might be the wrong decision, such as on the classic base-rate neglect task (Kahneman & Tversky, 1973). The crucial question is whether people know that they are focused on this bias. Recent dual process theories (DPTs) have incorporated the crucial role of conflict detection and resolution to better explain why people are biased on classic reasoning and judgment tasks. Two recent models, the Logical Intuition Model (De Neys, 2012) and the Three-stages Model (Pennycook, Fugelsang, & Koehler, 2015) suggest that the source of errors and bias can be explained in two distinct ways. The Logical Intuition Model postulates that people are generally efficient and routine conflict detectors and that errors are due to a failure to inhibit an initial, intuitive response. The Three-stages Model claims that detection is imperfect and that the detection mechanism is the main source of errors because people do not recognize that they are making biased decisions. These claims were investigated in a series of three experiments. In Experiment 1, participants completed a modified base-rate neglect task. In Experiment 2, a conditional reasoning task was added to test whether the claims of the two base-rate neglect models would transfer to a qualitatively different task. In Experiment 3, participants were placed into four different groups where they were given one of two forms of false feedback, true feedback, or no feedback (control) in order to test whether feedback would interact with answer confidence and resultant intuitive or analytic processing correlates (such as response time). Across all three experiments, the Three-stages Model’s claim that monitoring/detection failures is the main source of bias on the base-rate neglect task was supported over the Logical Intuition Model’s claim of inhibition failures as the major source of bias. Experiments 2 and 3 support the explanation that these two models are task-specific to base-rate neglect, as conditional reasoning behavioral patterns did not support either model fully. Feedback did not have the predicted effects on accuracy, response times, or confidence. Implications of these findings regarding general dual process theory, including the impact of methodological limitations, are discussed. Small modifications to the Three-stages Model are offered to reflect the data presented here.

Cognitive theories emphasise the role of attention and memory biases in the development and maintenance of body dissatisfaction, but there are critical gaps in the literature. Phase One of this thesis investigated the unexplored link between attention and memory for body-related information in a non-clinical sample. Study 1 found that the amount of visual attention allocated to body-related images was positively correlated with recognition memory for those images among both restrained and unrestrained eaters. Interestingly, restrained eaters had better recognition memory for body-related images than did unrestrained eaters regardless of the level of attention, suggesting that restrained eaters may be more efficient than unrestrained eaters at processing body-related information. The relationship between attention and memory remained even when participants were cognitively-distracted (Study 2), but was no longer observed when their visual attention was severely disrupted (Study 3). Study 4 further showed that both restrained and unrestrained eaters’ memory for body-related images was impaired by a schema-specific cognitive load. Overall, then, there is no evidence for the processing efficiency hypothesis. Phase Two investigated the relationship between memory for thin-ideal images and body dissatisfaction, as well as the mediational role of memory in the relationship between attention and body dissatisfaction. Attention was also manipulated by varying the exposure duration of thin-ideal images to demonstrate a causal relationship between attention and memory. Study 5 found that better recognition memory for thin-ideal images was associated with lower body dissatisfaction for both restrained and unrestrained eaters. However, Study 6 showed that, after one week, better recognition memory for thin-ideal images was associated with higher body dissatisfaction for restrained eaters. Memory mediated the relationship between attention to thin-ideal images and body dissatisfaction, both in the short-term (Study 5) and in the long-term (Study 6). Study 7 employed a sample of individuals with eating disorders and provided some tentative indication that the relationship between memory and body dissatisfaction may be generalisable to the clinical population. These findings highlight the important role of memory in influencing women’s body image, and suggest that existing treatments for body image issues may be enhanced if memory is also targeted.

Current explanations for anchoring phenomena seem to be unable to account for the diversity of effects found by 40 years of research. Additionally, the theories do not have much to say about the processes that make anchors so resilient to modification. I argue that by focusing on the mechanisms involved in spatial representation, we can account for most anchoring effects which have spatial components.

Information about environmental dangers is valuable and the cost of transmitting such information to social partners is minimal. Furthermore, an information transmitter who helps a social partner to avoid harm can later reap the benefits of a continued social relationship, and does not have to pay the costs of helping an injured friend. This cost-benefit assymetry suggests there is a positive selective pressure on a propensity to socially transmit information about danger at especially high rates compared to information that is not about danger. We call this predicted propensity a warning bias. Here we report the results of tests for this bias using data from the social networking site Twitter. Two coders rated each of 13,203 tweets (publicly-shared 140 character utterances) for whether or not each tweet contained information about danger. The number of retweets for each tweet indexes that tweet's transmission rate, and was our outcome variable. Results of negative binomial regressions showed that tweets about danger have up to 3.13 times as many retweets as tweets that are not about danger.

Cognitive biases have long been studied for their effects on an individual's decision-making. This study is the first of its kind to link these preferential thinking patterns to high school biology students' acceptance of evolution. At five urban high schools, 164 high school biology students were surveyed using a new instrument called the Cognitive Bias Assessment For Science Students – Evolution. The CBASS-E assesses both the level of non-acceptance of evolution, as well as how intensely participants demonstrate four specific cognitive biases. The four biases measured on the CBASS-E are anchoring, bandwagon effect, confirmation bias, and wishful thinking. This study showed that confirmation bias closely relates to a student's view of evolution. As part of this research, students were shown a video designed to teach high school students about cognitive bias and its effect on science learning. Results indicated that both cognitive bias and views on evolution are persistent, as student responses did not significantly change after watching the video. Included in this research are implications for teachers and administrators, as well as suggestions for further areas of study.

Public policy efforts to curb obesity often adhere to a rational actor model of human behavior, asserting that consumer behavior will change provided proper economic incentives, nutritional information, and health education. However, rigorous academic research related to such questions remains limited in scope and appears inconclusive as to the success of such economic and cognitive interventions. In contrast, research in social psychology and behavioral economics suggests that decision making is partially based on heuristics, or rules of thumb, and susceptible to powerful cognitive biases. External cues can subtly influence decision making in powerful ways. In this paper, after discussing existing policy efforts and their limitations, we use concepts from behavioral decision theory to design interventions related to different psychological domains in hopes of providing a more complete understanding of consumer dietary decision making. We move beyond traditional cognitive methods, namely the provision of nutritional information and economic incentives, to suggest the value of other cognitive, affective, social, and environmental influences in shaping food choices. Over a 21-month period, we tested 9 interventions in a point-of-purchase field study at a hospital cafeteria, focusing on the healthfulness of beverage purchases and chip purchases. Information, in the form of novel, reinforcing health messages, had the most consistently beneficial effect on the healthfulness of purchases. Traffic light colored-nutritional labeling, affect-based cues (smiley faces and frowny faces), and environmental changes including grouping items together based on level of healthfulness ("grouping by healthfulness") and pairing an unhealthy item with a healthier alternative ("healthy substitute pairing") also affected choices. Messages related to social norms had no effect on purchases. Our work adds to existing consumer behavior research and helps to inform health policy of additional cognitive factors and biases that must be taken into account when designing interventions and which can, indeed, be leveraged to influence dietary behavior. This is the first study of which we know to test the relative effects of this number and variety (economic, cognitive, affective, social, and environmental) of theory-based behavioral nudges on food choice in one setting.

Escalation of commitment is the voluntary continuation of investing resources into what appears to be a failing course of action whose outcome is uncertain. Investigation into the escalation of commitment phenomenon is important to organizations because such behavior could result in grave economic loss. This research investigates two cognitive biases that we posit lead to IT escalation of commitment, namely, attribution bias and overconfidence in an escalation decision, as well as desire to rectify past outcomes (DRPO) for its potential role as a mediator. To test our research model, 160 IT managers participated in a web-based role-playing experiment. Attribution was manipulated at two levels (internal and external), creating two treatment conditions. We posited that the participants assigned to the internal attribution condition would escalate their commitment to the failing IT project to a greater extent than participants assigned to the external attribution condition; that individuals that have a high, versus low, level of overconfidence would have a greater tendency to escalate; and that DRPO would mediate the effects of attribution and overconfidence on escalation of commitment. Attribution bias was significant at the .1 level, but in the opposite direction of what was hypothesized; overconfidence showed a significant main effect on escalation. The effect of attribution bias on escalation was significantly mediated by DRPO, but the effect of overconfidence on escalation was not mediated by DRPO. Implications of these findings for both research and practice are discussed.

Prior research has shown that people with depression tend to think more negatively than people without depression. However, there is disagreement in the literature as to whether people with the highest depressive symptom severity tend to make either overly negative judgments or accurate judgments. Empirical research addressing the differences between these hypotheses is needed because the theoretical basis for one of the most successful treatments for depression, cognitive therapy (CT), is based in the assumption that people with depression tend to think in an unrealistic, pessimistic manner. Treatment-seeking adults with major depressive disorder (N=67) participated in a series of prediction tasks before and after a 16-week course of CT. Healthy controls matched demographically to treatment completers (N=45) also completed these tasks. The tasks included future life event prediction, prediction of significant others’ ratings of participants’ personality characteristics, and prediction of daily affective experiences. Outcomes were recorded, and a measure of optimistic / pessimistic bias was calculated by using participants’ predictions and outcomes. Depressed clients were pessimistically biased on a portion of their predictions of daily affective experiences, optimistically biased on the life events task and other portions of the experience sampling task, and not systematically biased on the personality rating task. Though depressed clients’ predictions and outcomes at intake were more negative than those of controls, patients and controls largely did not significantly differ in their levels of bias. Additionally, though predictions and outcomes improved from intake to post-treatment among depressed clients, bias generally did not change. Overall, the study largely failed to identify evidence of bias among depressed clients or that bias was reduced with a course of cognitive therapy for depression. However, clients’ predictions did become more positive over the course of treatment. Further research should investigate potential moderators of pessimistic bias, such as presence of schemas in particular domains or use of specific coping strategies.

Cognitive bias and thought suppression are two maladaptive patterns of thinking that have been associated with borderline personality disorder (BPD). Negative cognitive biases related to BPD include thoughts that they are bad, powerless, or vulnerable and that the world is dangerous. Thought suppression is a maladaptive emotion regulation strategy where unwanted thoughts are intentionally pushed out of one’s consciousness. However, previous research has connected thought suppression and cognitive biases to BPD only via self-report measures. The present study examined whether a laboratory task meant to measure cognitive bias and thought suppression (Scrambled Sentences Test) would predict BPD features over and above self report measures of cognitive bias and thought suppression. A sample of 153 undergraduates completed self-report measures of BPD features, thought suppression, and negative cognitive biases, as well as the Scrambled Sentences Test (SST). Results showed that while the SST was a good predictor of cognitive biases, it did not predict thought suppression when self report measures were included. Recognizing the importance of negative cognitive bias in BPD may be useful in continued treatment development. Further research into other ways of measuring thought suppression and cognitive biases in the lab may be warranted.

This dissertation exploits variation in tax policies in order to better understand the incentive effects of taxation on individual and firm behavior. The results presented herein are intended to inform policy debates in the two areas examined--property and multinational taxation--as well as the formulation of tax policy more generally. Chapter II extends the standard model of property tax capitalization to a unique set of features of the Michigan property tax system in order to address a primitive question on taxpayer behavior: namely, do households understand the property tax implications of their home purchases? As shown, Michigan homebuyers fundamentally fail in their comprehension of the tax system, behaving--incorrectly--as though temporary idiosyncratic differences in property tax obligations would persist indefinitely. Beyond its stark distributional consequences and the efficiency implications of suboptimal housing consumption, cognitively-biased behavior of this nature may nevertheless mitigate welfare losses from the lock-in effect of acquisition-value based assessments which characterize nearly half of all states' property tax systems in the U.S. Michigan's implementation of such assessment limits may thereby conceivably constitute a step in the direction of optimal tax salience, with the findings of this chapter reinforcing the view that salience considerations ought to play an important role in tax policy design. Chapters III and IV adopt different methodologies to quantifying tax avoidance activity among U.S. multinational corporations resulting from differential reductions in domestic taxation of foreign earnings. In Chapter III, the case considered is that of the dividends received deduction (DRD) enacted under the American Jobs Creation Act of 2004, which provided temporary incentives for multinationals to reallocate domestic income to low-tax foreign jurisdictions in order to repatriate these immediately under preferential terms. Viewed as an experiment with 85 percent territoriality, estimates of a modest income reallocation response to the DRD may alleviate concerns associated with the tax avoidance consequences of implementing a territorial tax system. Chapter IV instead examines stock market reactions to a proposed DRD renewal and finds evidence consistent with investors rewarding multinational tax avoidance without offering the statistical precision necessary to make definitive conclusions as to its magnitude.

Humans have now amassed a sizable knowledge of widespread, nonconscious cognitive biases which affect our behavior, especially in social and economic contexts. I contend that a democratic government is uniquely justified in using knowledge of cognitive biases to promote pro-democratic behavior, conditionally justified in using it to accomplish ends traditionally within the scope of government authority, and unjustified in using it for any other purpose. I also contend that the government ought to redesign institutional infrastructure to avoid triggering cognitive biases where it is not permitted intentionally to manipulate such biases and to optimize the effects of such biases where permissible. I shall use the United States of America as an example throughout, but my conclusions apply equally to any democracy which values the political autonomy of its populace.

The present investigation tested the prediction that effortless control processing, the deliberate activation of innate automatic encoding mechanisms, will enable nondepressed persons to accurately judge degree of control. Subjective judgment of control in nondepressed students was examined by a modification of the method developed by Jenkins and Ward (1965). The modification was based on Hasher and Zacks' (1979) version of the method. Several measures were used to assess students' representations of control. Students were asked to judge the degree of control their responses had over outcomes rather than the degree of contingency between responses and outcomes. To facilitate comparison of prior studies on the judgment of contingency with the present study, Jenkins and Ward's (1965) index of the actual degree of control was used. Their index used the magnitude of the difference between the conditional probability of an outcome given the occurrence of one response versus the conditional probability of the outcome given the occurrence of another response as representing degree of control or contingency. In this experiment, students instructed in effortful control processing and effortless control processing were presented with a series of problems in which there was no contingency between their responses and outcomes. The problems differed in the degree of favorable outcome frequency. Students' abilities to detect noncontingency between responses and outcomes under different conditions of outcome frequency was examined.