Human Cognitive Biases Research Articles

Online and Stochastic Learning with a Human Cognitive Bias

Sequential learning for classification tasks is an effective tool in the machine learning community. In sequential learning settings, algorithms sometimes make incorrect predictions on data that were correctly classified in the past. This paper explicitly deals with such inconsistent prediction behavior. Our main contributions are 1) to experimentally show its effect for user utilities as a human cognitive bias, 2) to formalize a new framework by internalizing this bias into the optimization problem, 3) to develop new algorithms without memorization of the past prediction history, and 4) to show some theoretical guarantees of our derived algorithm for both online and stochastic learning settings. Our experimental results show the superiority of the derived algorithm for problems involving human cognition.

Leveraging position bias to improve peer recommendation.

With the advent of social media and peer production, the amount of new online content has grown dramatically. To identify interesting items in the vast stream of new content, providers must rely on peer recommendation to aggregate opinions of their many users. Due to human cognitive biases, the presentation order strongly affects how people allocate attention to the available content. Moreover, we can manipulate attention through the presentation order of items to change the way peer recommendation works. We experimentally evaluate this effect using Amazon Mechanical Turk. We find that different policies for ordering content can steer user attention so as to improve the outcomes of peer recommendation.

Five Ways to Incorporate Systems Thinking into Healthcare Organizations

Healthcare is a complex system that involves high risk to patients, clinicians, manufacturers, and other stakeholders. Clinicians process inordinate amounts of data and synthesize these inputs to make critical decisions that affect patient health and safety. The unprecedented advancement in medical technology during the past half-century has contributed substantially to healthcare’s expanding complexity.1 Consequently, clinicians have more knowledge to synthesize, tasks to perform, and processes to manage than ever before.1 The delivery of care often is overly complex and unstandardized, thereby reducing rather than increasing safety. For example, new devices are being implemented in medical practice at an astounding pace, requiring clinicians to continuously expand their breadth of knowledge and expertise. Consequently, a main cause of adverse events is misuse of medical devices. Variation in medical devices among hospitals (and within a given hospital) is a key cause of these adverse events. Design improvements and standardization of equipment have been suggested as ways to reduce errors. Although new technologies usually are advantageous for the patient, health professionals often encounter difficulties in using devices. Furthermore, current healthcare systems are not designed to support the care of complex patients, such as populations with chronic conditions requiring care management across multiple providers and services. In 2005, a joint report from the National Academy of Engineering and Institute of Medicine advocated the extensive application of systems thinking to improve the delivery of healthcare.2 Systems thinking centers on the dynamic interaction, synchronization, and integration of people, processes, and technology.3 By gaining an understanding of the dynamics among people, processes, and technology, systems thinking aids in recognizing how to intervene (e.g., focusing on changes to device design, clinician training, and/or clinical practice) in the system successfully. Moreover, systems thinking helps identify the critical relationships and connections often missed or undervalued that are pivotal to a successful implementation effort. If the high-risk healthcare industry focuses on an enhanced application of systems-thinking approaches, it will benefit from opportunities to reform the care delivery system by reducing unnecessary complexities and unexplained practice variations. This article describes five ways in which systems thinking can be incorporated into healthcare organizations. Rather than providing an exhaustive list of information, this report is meant to serve as a starting point for exploring how to integrate systems thinking.

The Argument from Silence

The argument from silence is a pattern of reasoning in which the failure of a known source to mention a particular fact or event is used as the ground of an inference, usually to the conclusion that the supposed fact is untrue or the supposed event did not actually happen. Such arguments are widely used in historical work, but they are also widely contested. This paper surveys some inadequate attempts to model this sort of argument, offers a new analysis using a Bayesian probabilistic framework that isolates the most problematic step in such arguments, illustrates a key problem besetting many uses of the argument, diagnoses the attraction of the argument in terms of a known human cognitive bias affecting the critical step, and suggests a standard that must be met in order for any argument from silence to have more than a very weak influence on historical reasoning.

The origins of duality of patterning in artificial whistled languages.

In human speech, a finite set of basic sounds is combined into a (potentially) unlimited set of well-formed morphemes. Hockett (1960) placed this phenomenon under the term 'duality of patterning' and included it as one of the basic design features of human language. Of the thirteen basic design features Hockett proposed, duality of patterning is the least studied and it is still unclear how it evolved in language. Recent work shedding light on this is summarized in this paper and experimental data is presented. This data shows that combinatorial structure can emerge in an artificial whistled language through cultural transmission as an adaptation to human cognitive biases and learning. In this work the method of experimental iterated learning (Kirby et al. 2008) is used, in which a participant is trained on the reproductions of the utterances the previous participant learned. Participants learn and recall a system of sounds that are produced with a slide whistle. Transmission from participant to participant causes the whistle systems to change and become more learnable and more structured. These findings follow from qualitative observations, quantitative measures and a follow-up experiment that tests how well participants can learn the emerged whistled languages by generalizing from a few examples.

Human Preferences for Symmetry: Subjective Experience, Cognitive Conflict and Cortical Brain Activity

This study examines the links between human perceptions, cognitive biases and neural processing of symmetrical stimuli. While preferences for symmetry have largely been examined in the context of disorders such as obsessive-compulsive disorder and autism spectrum disorders, we examine various these phenomena in non-clinical subjects and suggest that such preferences are distributed throughout the typical population as part of our cognitive and neural architecture. In Experiment 1, 82 young adults reported on the frequency of their obsessive-compulsive spectrum behaviors. Subjects also performed an emotional Stroop or variant of an Implicit Association Task (the OC-CIT) developed to assess cognitive biases for symmetry. Data not only reveal that subjects evidence a cognitive conflict when asked to match images of positive affect with asymmetrical stimuli, and disgust with symmetry, but also that their slowed reaction times when asked to do so were predicted by reports of OC behavior, particularly checking behavior. In Experiment 2, 26 participants were administered an oddball Event-Related Potential task specifically designed to assess sensitivity to symmetry as well as the OC-CIT. These data revealed that reaction times on the OC-CIT were strongly predicted by frontal electrode sites indicating faster processing of an asymmetrical stimulus (unparallel lines) relative to a symmetrical stimulus (parallel lines). The results point to an overall cognitive bias linking disgust with asymmetry and suggest that such cognitive biases are reflected in neural responses to symmetrical/asymmetrical stimuli.

Toward a Neurobiology of Unrealistic Optimism

Toward a Neurobiology of Unrealistic Optimism

Meta-Analysis Using Individual Participant Data Is the Gold Standard for Diagnostic Studies Δ

We highly appreciate the comments by Zheng et al. on the use of individual participant data (IPD) in diagnostic studies. We agree that meta-analysis of using IPD from multiple clinical studies enables detailed investigation of diagnostic studies. We also found that increasing numbers of IPD meta-analyses (IPDMA) from observational data are being conducted to enhance the statistical power and detail of epidemiological studies,1 and this approach is being increasingly applied in many research studies.2 However, there is some skepticism as to whether IPDMA is really the gold standard for diagnostic studies, considering the fact that it involves a number of challenges. On this point, we have our own view. First, in order to get a better comparison with the previous study conducted by Shaheen and Myers in 2007,3 we used aggregate data (AD) instead of IPD to perform this updated meta-analysis. Second, extra cost in effort, time, and complexity is required to obtain and manage raw data in IPDMA. Commendable examples of IPDMA are those conducted by the Emerging Risk Factors Collaboration (ERFC),4 who have remarkably collected IPD from 116 prospective studies and more than 1.2 million participants. Although meta-analysis methods using AD are well established and fairly routine, methods for IPDMA are more complex but less well known. Currently, although the number of published articles related to “meta-analysis” has risen dramatically from hundreds of articles per year in the early 1990s to thousands of articles every year since 2003, the number of articles of IPDMA only accounts for a negligible proportion of less than 1.71% (Fig. 1). Number of articles related “meta-analyses” published in PubMed from January 1991 to December 25, 2010 compared with number of distinct, applied IPDMA published from January 1991 to December 25, 2010, as identified by a systematic review of PubMed. Finally, should one embark on an IPDMA when few studies provide their IPD, making it difficult to estimate random effects? Likewise, is an IPDMA reliable when only a proportion of existing studies provide IPD? Unfortunately this is the current situation that regrettably leads to what Riley5 referred to as availability bias—a human cognitive bias that tends to overestimate probabilities of events associated with memorable or vivid occurrences, where studies that provide IPD are a kind of biased subset of all existing studies. In conclusion, we think that methods for IPDMA are prone to be affected by bias with inadequate generalizability despite their widely recognized strength. The appropriate strategy at this moment is probably to use both approaches in a complementary fashion, in which an AD meta-analysis is conducted in the first step rather than an IPDMA. As Riley pointed out, IPD is not the be-all and end-all for meta-analysis just yet.5 Yong-Ning Xin M.D.* , Zhong-Hua Lin M.D. , An-Jin Chen M.D.* , Shi-Ying Xuan M.D.* , * College of Medicine and Pharmaceutics Ocean University of China Qingdao, Shandong Province, China, Qingdao Municipal Hospital, Qingdao Shandong Province, China, School of Medicine, Qingdao University, Qingdao Shandong Province, China.

5.3.2 Risk Cube Data Biases

AbstractRisk cubes, actually a 5times5 matrix, used in industry characterize particular risks in terms of the likelihood of occurrence, and the consequence of the actualized risk. Human cognitive bias research led by Daniel Kahneman and Amos Tversky exposed systematic translations of objective probability and value as judged by human subjects. An examination of industry‐generated risk cube data reveals evidence of biases in the judgment of likelihood and consequence. In addition, other biases – including a ‘diagonal bias’ – are revealed in the risk cube data. The evidence presented could improve risk cube based risk analysis.

Home Loan Exit Fees: The Cost of Ending a Home Loan Early

The fees charged by financial institutions are coming under increasing scrutiny. In this paper exit fees, or the fees charged by financial institutions for early termination of a variable rate mortgage, are analysed. The authors discuss why these fees are a concern, including the potential impact on competition in the home loan market, the high rate of refinancing, and the way in which exit fees can exploit human cognitive biases. The authors also review the regulation of exit fees and present data on exit fees drawn from a study of 198 home loans offered by a wide range of financial institutions and also data obtained from two consumer law centres. The authors conclude with observations relating to the new role of the Australian Securities and Investments Commission in the regulation of exit fees.

Risk matrix input data biases

Risk matrices used in industry characterize particular risks in terms of the likelihood of occurreRisk matrix input data biaseshe actualized risk. Human cognitive bias research led by Daniel Kahnem...

Choosing appropriate teaching and training techniques

This paper describes an approach to teaching and training techniques based on types of human memory and skill and on cognitive biases. The focus is on low level 'craft skill' techniques. The first part of the paper shows how a framework of explicit, semi-tacit and tacit types of memory and skill can provide guidance on choice of appropriate teaching and training techniques. The second part describes how training and teaching can turn human cognitive biases to advantage by presenting material in ways which make it more accessible and memorable. We also briefly describe some of the underlying neurophysiological issues involved.

Risk matrix input data biases

AbstractRisk matrices used in industry characterize particular risks in terms of the likelihood of occurreRisk matrix input data biaseshe actualized risk. Human cognitive bias research led by Daniel Kahneman and Amos Tversky exposed systematic translations of objective probability and value as judged by human subjects. Applying these translations to the risk matrix allows the formation of statistical hypotheses of risk point placement biases. Industry‐generated risk matrix data reveals evidence of biases in the judgment of likelihood and consequence—principally, likelihood centering, a systematic increase in consequence, and a diagonal bias. Statistical analyses are conducted with linear regression, normal distribution fitting, and Bayesian analysis. Evidence presented could improve risk matrix based risk analysis prevalent in industry. © 2008 Wiley Periodicals, Inc. Syst Eng

KR17 Risk Cube Data Biases

AbstractRisk cubes used in industry characterize particular risks in terms of the likelihood of occurrence, and the consequence of the actualized risk. Human cognitive bias research led by Daniel Kahneman and Amos Tversky exposed systematic translations of objective probability and value as judged by human subjects. An examination of industry‐generated risk cube data reveals evidence of biases in the judgment of likelihood and consequence. In addition, other biases – including a ‘diagonal bias’ – are revealed in the risk cube data. The evidence presented could improve risk cube based risk analysis.

The impact of health communication on health‐related decision making

PurposeThe purpose of this paper is to review evidence related to the factors that influence people's understanding of health information and how miscommunication of health information can jeopardize people's health.Design/methodology/approachA literature review was conducted of English language articles, cited in major literature databases from the last 40 years, which describe factors related to comprehension of health information. A total of 93 articles were included.FindingsThe paper finds that health communication should take into consideration the role of the following factors on the processing and interpretation of health information: health literacy, format presentation of information, and human cognitive biases and affective/personal influences.Practical implicationsHealth communication is a major component of health care. Every health care encounter involves exchange of information, which is intended to enhance people's knowledge in order to assist them to make an informed decision about their health care. However, the mere act of providing information does not guarantee comprehension. People's comprehension of information depends on several factors, including health literacy and numeracy skills, the format presentation of health information and human cognitive biases in the information processing and interpretation. Ineffective health communication can result in a wide range of direct and indirect health consequences including failure to understand and comply with treatment, poorer health status, increased risk of injuries, increased hospitalization, and decreased use of preventive services.Originality/valueThis paper provides health professionals and educators with an overview of important issues related to health communication and highlights strategies that facilitate effective communication to help people to make informed decisions about their health care.

Neural correlates of human cognitive bias resulting in illogical inference

Neural correlates of human cognitive bias resulting in illogical inference

6.2.1 Tradeoff Studies and Cognitive Biases

AbstractTradeoff studies involving human subjective calibration and data updating are often distrusted by decision makers. A review of objectivity and subjectivity in decision making shows that Prospect Theory is a good model for actual human decision making. Similarities between the elements of tradeoff studies and the elements of experiments in cognitive science show that tradeoff studies are susceptible to human cognitive biases. Examples of relevant biases are given.

Reappraisal of Decision-Making Models in Engineering Applications

Because of substantial advances of automatic technology in human-machine systems, the operator has changed to a maker in a control room. Modeling operator's decision behavior has attracted a major interest in research fields such as human factors and reliability engineering. However, traditional reasoning models proposed by psychologists are so that they are deficient in accounting for the interactive nature of modern human-machine systems. Concerning this issue, a conceptual framework of decision making is suggested to distinguish between dynamic and static decision making. Hence, dynamic and static decision processes are defined considering the decision stages in a decision-making task. Reasoning is regarded as a dominant mental activity in each decision stage and the inadequacy of static reasoning models in handling dynamic decision making is discussed. Considering human cognitive biases and the requirements of engineering applications, an event-driven approach, called the Dynamic B...

Normative-Descriptive Modeling of Human Teams: A 15-Year Perspective

For almost 15 years we have been conducting combined analytical modeling and empirical research in the area of distributed decisionmaking and coordination in multi-person teams. Our emphasis has been on the coordination of information, resources and actions within the team. Using a unique team-in-the-loop simulation environment, experiments have been designed and operationalized to examine team decisionmaking and coordination processes, and to manipulate those aspects of the problem that affect these processes. Our concomitant modeling efforts have been geared to establish a mathematical framework for explaining and predicting team performance and strategy. Our approach blends normative mathematical theories for team decisionmaking with descriptive facets that capture known, and observed, human cognitive limitations and biases. The resulting normative-descriptive models have provided excellent predictions of the actual data collected in the experiments, and have shed new light on the nature of team coordination processes.

Individual learning of coordination knowledge

. Social agents, both human and computational, inhabiting a world containing multiple active agents, need to coordinate their activities. This is because agents share resources, and without proper coordination or ‘rules of the road’, everybody will be interfering with the plans of others. As such, we need coordination schemes that allow agents to effectively achieve local goals without adversely affecting the problem-solving capabilities of other agents. Researchers in the field of Distributed Artificial Intelligence (DAI) have developed a variety of coordination schemes under different assumptions about agent capabilities and relationships. Whereas some of these researchers have been motivated by human cognitive biases, others have approached it as an engineering problem of designing the most effective coordination architecture or protocol. We evaluate individual and concurrent learning by multiple, autonomous agents as a means for acquiring coordination knowledge. We show that a uniform reinforcement learning algorithm suffices as a coordination mechanism in both cooperative and adversarial situations. Using a number of multi-agent learning scenarios with both tight and loose coupling between agents and with immediate as well as delayed feedback, we demonstrate that agents can consistently develop effective policies to coordinate their actions without explicit information sharing. We demonstrate the viabilityof using both the Q-learning algorithm and genetic algorithm based classifier systems with different pay-off schemes, namely the bucket brigade algorithm (BBA) and the profit sharing plan (PSP), for developing agent coordination on two different multi-agent domains. In addition, we show that a semi-random scheme for action selection is preferable to the more traditional fitness proportionate selection scheme used in classifier systems.