DOCSLIB.ORG

An Analytics Approach to Debiasing Asset- Management Decisions

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
An Analytics Approach to Debiasing Asset- Management Decisions An analytics approach to debiasing asset- management decisions Wealth & Asset Management December 2017 Nick Hoffman Martin Huber Magdalena Smith An analytics approach to debiasing asset-management decisions Asset managers can find a competitive edge in debiasing techniques— accelerated with advanced analytics. Investment managers are pursuing analytics- Richard H. Thaler, one of the leaders in this field aided improvements in many different areas of (see sidebar, “The debiasing nudge”). the business, from customer and asset acquisition to operations and overhead costs. The change Asset management: An industry facing area we focus on in this discussion is investment challenges performance improvement, specifically the Investment managers are facing significant debiasing of investment decisions. With the challenges to profitability. Demonstrating the help of more advanced analytics than they are value of active management has become more already using, funds have been able to measure difficult in a market where returns are narrowing. the role played by bias in suboptimal trading Dissatisfaction with active performance is causing decisions, connecting particular biases customers to migrate toward cheaper passive to particular decisions. Such discoveries funds (Exhibit 1). Actions active funds are taking in provide the necessary foundation for effective response include reducing fees, which has created countermeasures—the debiasing methods that can a competitive cycle that is compressing margins. bring significant performance improvements to a Some funds (such as Allianz Global Investors pressured industry. and Fidelity) are changing their fee and pricing structures to make them more dependent on Business leaders are increasingly recognizing the outperformance through the use of fulcrum-type risk of bias in business decision making. Insights fee structures in their funds. from the fields of behavioral economics and cognitive psychology continually emerge to reveal To push back on these trends, some investment that individuals and institutions do not base managers are broaching the topic of bias in financial and other decisions on purely rational investment decisions. They are seeking a considerations. The contours of the irrational competitive edge and—perhaps more important— biases have become increasingly known, and the to improve the value proposition of active ways bias operates in our thought processes can management. Evidence from within and outside often be predicted. Most important of all, the the industry strongly suggests that even the leading methods and means to counteract biases are asset managers in top-performing funds could becoming more sophisticated and effective, at least improve their investment performance by applying for those executives willing to inquire into their debiasing techniques. Our recent experience debiasing needs. working with investment managers has shown that enhancing these techniques with analytics can Advances in the use of debiasing techniques to improve performance significantly. improve decision making have been inspired by the research of many pioneering scholars. Bias and debiasing in action Their innovations have had numerous practical Bias is a risk in all business decision making— confirmations in the public sector and business the more significant the decision, the greater the settings as well as in the decisions of individuals. risk. Particular biases have been behind many As if in recognition of the deepening relevance costly missteps by companies and institutions in of debiasing in economic life, the Nobel Prize every sector. Consequently, behavioral scientists in Economic Sciences was awarded this year to and business leaders have developed methods for 2 An analytics approach to debiasing asset-management decisions The debiasing nudge In October 2017, the Nobel Committee awarded its decision making is undermined by biases. Among prize in Economic Sciences to Richard H. Thaler of the biases the authors analyzed are the stability the University of Chicago, who has thought deeply biases to which investment decision making is highly about rational and irrational behaviors in economic susceptible: anchoring, for example—the tying decision making. He and his colleagues in the field of of actions to an initial value and failure to adjust to behavioral economics—including Dan Ariely, Amos take into account new information. Loss aversion Tversky, and Daniel Kahneman (a previous Nobel is another such bias—the familiar fear that makes winner)—analyze the psychological dimension of us more risk averse than logic would allow. Where economics, partly leaning on insights from the field of human decision making is more prone to bias than cognitive psychology. Their research reveals patterns to reasonable deliberation, Thaler and Sunstein that did not align with the rational assumptions recommend the debiasing “nudge”—a benign, often embedded in prevailing descriptions of economic small adjustment that counters irrational impulses. systems and individual financial actions. They The authors discuss many examples of successful found that especially under conditions of risk and nudges; reviews of Nudge most often cited the opt- uncertainty, individuals as well as institutions fail to in default for employee retirement savings plans. In behave as might be expected when making financial this example, many more employees will save for decisions. Assumed universal principles, such as retirement when they are automatically enrolled in actual self-interest and a sure grasp of dynamic a plan. That is, the nudge of requiring employees inputs including time and probability, often give way to to save unless they opt out of the plan gets better irrational and unpredictable actions, based on narrow results than an approach that requires them to opt in. or flawed data and personal experiences. Everyone recognizes the need to save for retirement, but not everyone acts on it. The opt-in nudge is in fact In their book, Nudge: Improving Decisions About a form of debiasing for those irrationally ignoring their Health, Wealth, and Happiness (2008), Thaler and own future financial well-being. coauthor Cass Sunstein reveal ways that rational debiasing decision making. Many of the companies top performers were those that least anchored that have adopted these methods can attest to their their capital-allocation decisions to decisions effectiveness in improving outcomes: made the previous year. “Dynamic reallocators” achieved a median return that was nearly McKinsey cross-sector research has suggested four percentage points higher than that of that one of the most prevalent irrational biases, “Dormant reallocators.” the form of stability bias known as anchoring, undermines optimal capital reallocation. The An executive at the German electric utility research compared capital allocation and total RWE recently discussed his experience leading shareholder returns over a 20-year period a debiasing effort.1 Like most of its peers, the (1990 to 2010) at more than 1,500 publicly company had until recently based capital- traded companies in the United States. The investment decisions on ever-rising commodity An analytics approach to debiasing asset-management decisions 3 CDP 2017 An analytics approach to debiasing asset-management decisions Exhibit 1 of 4 Exhibit 1 Customers are migrating to passive equity funds, which have lower margins. North America net ow growth and revenue margin by asset class1 Passive Active Alternatives Bubble size 2016 AUM 150 145 Public market alternatives 60 Active specialty 55 equity Active specialty 50 Multi-asset fixed income basis points 45 40 35 Active core 30 Active core equity fixed income 25 Passive 20 Passive fixed 15 equity Passive Passive income 10 Money other 2016 revenue margin, multi-asset 5 market 0 –13 –12 –11 –10 –9 –8 –7 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 2016 net ows, % 1 Active core equity includes US large cap and yieldincome equity; active core fixed income includes core, core plus, and municipal bonds; active specialty equity includes foreign, global, EM and US smallmid-cap; active specialty fixed income includes global, EM, high yield, TIPS, and unconstrained. Source: 2016 McKinsey Performance Lens Global Asset Management Survey and Global Growth Cube and power prices. RWE leaders realized that A global investment bank reviewed its traders’ debiasing efforts could have challenged their decisions on equity position weighting. The earlier assumptions, enabling them to hedge bank recognized overcommitment to positions potential adverse effects. For future decisions, and suboptimal execution due to endowment they implemented a farsighted cultural change effects and confirmation bias. The endowment program to identify and counter cognitive effect, discussed in the research of Richard biases throughout the organization. Care was Thaler and other behavioral economists, refers taken to ensure that dissenting and outside to the psychological effects that prejudice analyses were fully articulated. For important owners in favor of retaining their assets despite moves, a “devil’s advocate” was appointed; for changing conditions. Confirmation bias is a the company’s most pressing strategic problem, form of pattern-recognition bias, causing us separate internal and external teams—a to see nonexistent patterns in information: “red team–blue team” approach—were assigned evidence supporting a favored belief is to develop solutions independently of each overvalued,
Load more

Recommended publications
  • The Status Quo Bias and Decisions to Withdraw Life-Sustaining Treatment
    HUMANITIES | MEDICINE AND SOCIETY The status quo bias and decisions to withdraw life-sustaining treatment n Cite as: CMAJ 2018 March 5;190:E265-7. doi: 10.1503/cmaj.171005 t’s not uncommon for physicians and impasse. One factor that hasn’t been host of psychological phenomena that surrogate decision-makers to disagree studied yet is the role that cognitive cause people to make irrational deci- about life-sustaining treatment for biases might play in surrogate decision- sions, referred to as “cognitive biases.” Iincapacitated patients. Several studies making regarding withdrawal of life- One cognitive bias that is particularly show physicians perceive that nonbenefi- sustaining treatment. Understanding the worth exploring in the context of surrogate cial treatment is provided quite frequently role that these biases might play may decisions regarding life-sustaining treat- in their intensive care units. Palda and col- help improve communication between ment is the status quo bias. This bias, a leagues,1 for example, found that 87% of clinicians and surrogates when these con- decision-maker’s preference for the cur- physicians believed that futile treatment flicts arise. rent state of affairs,3 has been shown to had been provided in their ICU within the influence decision-making in a wide array previous year. (The authors in this study Status quo bias of contexts. For example, it has been cited equated “futile” with “nonbeneficial,” The classic model of human decision- as a mechanism to explain patient inertia defined as a treatment “that offers no rea- making is the rational choice or “rational (why patients have difficulty changing sonable hope of recovery or improvement, actor” model, the view that human beings their behaviour to improve their health), or because the patient is permanently will choose the option that has the best low organ-donation rates, low retirement- unable to experience any benefit.”) chance of satisfying their preferences.
    [Show full text]
  • Avoiding the Conjunction Fallacy: Who Can Take a Hint?
    Avoiding the conjunction fallacy: Who can take a hint? Simon Klein Spring 2017 Master’s thesis, 30 ECTS Master’s programme in Cognitive Science, 120 ECTS Supervisor: Linnea Karlsson Wirebring Acknowledgments: The author would like to thank the participants for enduring the test session with challenging questions and thereby making the study possible, his supervisor Linnea Karlsson Wirebring for invaluable guidance and good questions during the thesis work, and his fiancée Amanda Arnö for much needed mental support during the entire process. 2 AVOIDING THE CONJUNCTION FALLACY: WHO CAN TAKE A HINT? Simon Klein Humans repeatedly commit the so called “conjunction fallacy”, erroneously judging the probability of two events occurring together as higher than the probability of one of the events. Certain hints have been shown to mitigate this tendency. The present thesis investigated the relations between three psychological factors and performance on conjunction tasks after reading such a hint. The factors represent the understanding of probability and statistics (statistical numeracy), the ability to resist intuitive but incorrect conclusions (cognitive reflection), and the willingness to engage in, and enjoyment of, analytical thinking (need-for-cognition). Participants (n = 50) answered 30 short conjunction tasks and three psychological scales. A bimodal response distribution motivated dichotomization of performance scores. Need-for-cognition was significantly, positively correlated with performance, while numeracy and cognitive reflection were not. The results suggest that the willingness to engage in, and enjoyment of, analytical thinking plays an important role for the capacity to avoid the conjunction fallacy after taking a hint. The hint further seems to neutralize differences in performance otherwise predicted by statistical numeracy and cognitive reflection.
    [Show full text]
  • Cognitive Bias Mitigation: How to Make Decision-Making More Rational?
    Cognitive Bias Mitigation: How to make decision-making more rational? Abstract Cognitive biases distort judgement and adversely impact decision-making, which results in economic inefficiencies. Initial attempts to mitigate these biases met with little success. However, recent studies which used computer games and educational videos to train people to avoid biases (Clegg et al., 2014; Morewedge et al., 2015) showed that this form of training reduced selected cognitive biases by 30 %. In this work I report results of an experiment which investigated the debiasing effects of training on confirmation bias. The debiasing training took the form of a short video which contained information about confirmation bias, its impact on judgement, and mitigation strategies. The results show that participants exhibited confirmation bias both in the selection and processing of information, and that debiasing training effectively decreased the level of confirmation bias by 33 % at the 5% significance level. Key words: Behavioural economics, cognitive bias, confirmation bias, cognitive bias mitigation, confirmation bias mitigation, debiasing JEL classification: D03, D81, Y80 1 Introduction Empirical research has documented a panoply of cognitive biases which impair human judgement and make people depart systematically from models of rational behaviour (Gilovich et al., 2002; Kahneman, 2011; Kahneman & Tversky, 1979; Pohl, 2004). Besides distorted decision-making and judgement in the areas of medicine, law, and military (Nickerson, 1998), cognitive biases can also lead to economic inefficiencies. Slovic et al. (1977) point out how they distort insurance purchases, Hyman Minsky (1982) partly blames psychological factors for economic cycles. Shefrin (2010) argues that confirmation bias and some other cognitive biases were among the significant factors leading to the global financial crisis which broke out in 2008.
    [Show full text]
  • Drug Exposure Definition and Healthy User Bias Impacts on the Evaluation of Oral Anti Hyperglycemic Therapies
    Drug Exposure Definition and Healthy User Bias Impacts on the Evaluation of Oral Anti Hyperglycemic Therapies by Maxim Eskin A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Health Policy Research School of Public Health University of Alberta © Maxim Eskin, 2017 ABSTRACT Accurate estimation of medication use is an essential component of any pharmacoepidemiological research as exposure misclassification will threaten study validity and lead to spurious associations. Many pharmacoepidemiological studies use simple definitions, such as the categorical “any versus no use” to classify oral antihyperglycemic medications exposure, which has potentially serious drawbacks. This approach has led to numerous highly publicized observational studies of metformin effect on health outcomes reporting exaggerated relationships that were later contradicted by randomized controlled trials. Although selection bias, unmeasured confounding, and many other factors contribute to the discrepancies, one critical element, which is often overlooked, is the method used to define exposure. Another factor that may provide additional explanation for the discrepancy between randomized controlled trials and observational study results of the association between metformin use and various health outcomes is the healthy user effect. Aspects of a healthy lifestyle such as following a balanced diet, exercising regularly, avoiding tobacco use, etc., are not recorded in typical administrative databases and failure to account for these factors in observational studies may introduce bias and produce spurious associations. The influence of a healthy user bias has not been fully examined when evaluating the effect of oral antihyperglycemic therapies in observational studies. It is possible that some, if not all, of the benefit, observed with metformin in observational studies may be related to analytic design and exposure definitions.
    [Show full text]
  • Graphical Techniques in Debiasing: an Exploratory Study
    GRAPHICAL TECHNIQUES IN DEBIASING: AN EXPLORATORY STUDY by S. Bhasker Information Systems Department Leonard N. Stern School of Business New York University New York, New York 10006 and A. Kumaraswamy Management Department Leonard N. Stern School of Business New York University New York, NY 10006 October, 1990 Center for Research on Information Systems Information Systems Department Leonard N. Stern School of Business New York University Working Paper Series STERN IS-90-19 Forthcoming in the Proceedings of the 1991 Hawaii International Conference on System Sciences Center for Digital Economy Research Stem School of Business IVorking Paper IS-90-19 Center for Digital Economy Research Stem School of Business IVorking Paper IS-90-19 2 Abstract Base rate and conjunction fallacies are consistent biases that influence decision making involving probability judgments. We develop simple graphical techniques and test their eflcacy in correcting for these biases. Preliminary results suggest that graphical techniques help to overcome these biases and improve decision making. We examine the implications of incorporating these simple techniques in Executive Information Systems. Introduction Today, senior executives operate in highly uncertain environments. They have to collect, process and analyze a deluge of information - most of it ambiguous. But, their limited information acquiring and processing capabilities constrain them in this task [25]. Increasingly, executives rely on executive information/support systems for various purposes like strategic scanning of their business environments, internal monitoring of their businesses, analysis of data available from various internal and external sources, and communications [5,19,32]. However, executive information systems are, at best, support tools. Executives still rely on their mental or cognitive models of their businesses and environments and develop heuristics to simplify decision problems [10,16,25].
    [Show full text]
  • Cognitive Biases in Economic Decisions – Three Essays on the Impact of Debiasing
    TECHNISCHE UNIVERSITÄT MÜNCHEN Lehrstuhl für Betriebswirtschaftslehre – Strategie und Organisation Univ.-Prof. Dr. Isabell M. Welpe Cognitive biases in economic decisions – three essays on the impact of debiasing Christoph Martin Gerald Döbrich Abdruck der von der Fakultät für Wirtschaftswissenschaften der Technischen Universität München zur Erlangung des akademischen Grades eines Doktors der Wirtschaftswissenschaften (Dr. rer. pol.) genehmigten Dissertation. Vorsitzender: Univ.-Prof. Dr. Gunther Friedl Prüfer der Dissertation: 1. Univ.-Prof. Dr. Isabell M. Welpe 2. Univ.-Prof. Dr. Dr. Holger Patzelt Die Dissertation wurde am 28.11.2012 bei der Technischen Universität München eingereicht und durch die Fakultät für Wirtschaftswissenschaften am 15.12.2012 angenommen. Acknowledgments II Acknowledgments Numerous people have contributed to the development and successful completion of this dissertation. First of all, I would like to thank my supervisor Prof. Dr. Isabell M. Welpe for her continuous support, all the constructive discussions, and her enthusiasm concerning my dissertation project. Her challenging questions and new ideas always helped me to improve my work. My sincere thanks also go to Prof. Dr. Matthias Spörrle for his continuous support of my work and his valuable feedback for the articles building this dissertation. Moreover, I am grateful to Prof. Dr. Dr. Holger Patzelt for acting as the second advisor for this thesis and Professor Dr. Gunther Friedl for leading the examination board. This dissertation would not have been possible without the financial support of the Elite Network of Bavaria. I am very thankful for the financial support over two years which allowed me to pursue my studies in a focused and efficient manner. Many colleagues at the Chair for Strategy and Organization of Technische Universität München have supported me during the completion of this thesis.
    [Show full text]
  • Working Memory, Cognitive Miserliness and Logic As Predictors of Performance on the Cognitive Reflection Test
    Working Memory, Cognitive Miserliness and Logic as Predictors of Performance on the Cognitive Reflection Test Edward J. N. Stupple ([email protected]) Centre for Psychological Research, University of Derby Kedleston Road, Derby. DE22 1GB Maggie Gale ([email protected]) Centre for Psychological Research, University of Derby Kedleston Road, Derby. DE22 1GB Christopher R. Richmond ([email protected]) Centre for Psychological Research, University of Derby Kedleston Road, Derby. DE22 1GB Abstract Most participants respond that the answer is 10 cents; however, a slower and more analytic approach to the The Cognitive Reflection Test (CRT) was devised to measure problem reveals the correct answer to be 5 cents. the inhibition of heuristic responses to favour analytic ones. The CRT has been a spectacular success, attracting more Toplak, West and Stanovich (2011) demonstrated that the than 100 citations in 2012 alone (Scopus). This may be in CRT was a powerful predictor of heuristics and biases task part due to the ease of administration; with only three items performance - proposing it as a metric of the cognitive miserliness central to dual process theories of thinking. This and no requirement for expensive equipment, the practical thesis was examined using reasoning response-times, advantages are considerable. There have, moreover, been normative responses from two reasoning tasks and working numerous correlates of the CRT demonstrated, from a wide memory capacity (WMC) to predict individual differences in range of tasks in the heuristics and biases literature (Toplak performance on the CRT. These data offered limited support et al., 2011) to risk aversion and SAT scores (Frederick, for the view of miserliness as the primary factor in the CRT.
    [Show full text]
  • 10. Sample Bias, Bias of Selection and Double-Blind
    SEC 4 Page 1 of 8 10. SAMPLE BIAS, BIAS OF SELECTION AND DOUBLE-BLIND 10.1 SAMPLE BIAS: In statistics, sampling bias is a bias in which a sample is collected in such a way that some members of the intended population are less likely to be included than others. It results in abiased sample, a non-random sample[1] of a population (or non-human factors) in which all individuals, or instances, were not equally likely to have been selected.[2] If this is not accounted for, results can be erroneously attributed to the phenomenon under study rather than to the method of sampling. Medical sources sometimes refer to sampling bias as ascertainment bias.[3][4] Ascertainment bias has basically the same definition,[5][6] but is still sometimes classified as a separate type of bias Types of sampling bias Selection from a specific real area. For example, a survey of high school students to measure teenage use of illegal drugs will be a biased sample because it does not include home-schooled students or dropouts. A sample is also biased if certain members are underrepresented or overrepresented relative to others in the population. For example, a "man on the street" interview which selects people who walk by a certain location is going to have an overrepresentation of healthy individuals who are more likely to be out of the home than individuals with a chronic illness. This may be an extreme form of biased sampling, because certain members of the population are totally excluded from the sample (that is, they have zero probability of being selected).
    [Show full text]
  • The Challenge of Debiasing Personnel Decisions: Avoiding Both Under- and Overcorrection
    University of Pennsylvania ScholarlyCommons Management Papers Wharton Faculty Research 12-2008 The Challenge of Debiasing Personnel Decisions: Avoiding Both Under- and Overcorrection Philip E. Tetlock University of Pennsylvania Gregory Mitchell Terry L. Murray Follow this and additional works at: https://repository.upenn.edu/mgmt_papers Part of the Business Administration, Management, and Operations Commons Recommended Citation Tetlock, P. E., Mitchell, G., & Murray, T. L. (2008). The Challenge of Debiasing Personnel Decisions: Avoiding Both Under- and Overcorrection. Industrial and Organizational Psychology, 1 (4), 439-443. http://dx.doi.org/10.1111/j.1754-9434.2008.00084.x This paper is posted at ScholarlyCommons. https://repository.upenn.edu/mgmt_papers/32 For more information, please contact [email protected]. The Challenge of Debiasing Personnel Decisions: Avoiding Both Under- and Overcorrection Disciplines Business Administration, Management, and Operations This journal article is available at ScholarlyCommons: https://repository.upenn.edu/mgmt_papers/32 1 The Challenge of Debiasing Personnel Decisions: Avoiding Both Under- and Over-Correction Philip E. Tetlock, * Gregory Mitchell, ** and Terry L. Murray *** Introduction This commentary advances two interrelated scientific arguments. First, we endorse Landy's (2008) concerns about the insufficient emphasis placed on individuating information by scholars eager to import social-cognition work on stereotyping into employment law. Building on Landy’s analysis, we emphasize that greater attention needs to be given to the power of accountability and teamwork incentives to motivate personnel decision-makers to seek and utilize individuating information that is predictive of job-relevant behavior. Second, we note how easy it is for exchanges between proponents and skeptics of unconscious stereotyping to lead to ideological stalemates.
    [Show full text]
  • Observational Studies and Bias in Epidemiology
    The Young Epidemiology Scholars Program (YES) is supported by The Robert Wood Johnson Foundation and administered by the College Board. Observational Studies and Bias in Epidemiology Manuel Bayona Department of Epidemiology School of Public Health University of North Texas Fort Worth, Texas and Chris Olsen Mathematics Department George Washington High School Cedar Rapids, Iowa Observational Studies and Bias in Epidemiology Contents Lesson Plan . 3 The Logic of Inference in Science . 8 The Logic of Observational Studies and the Problem of Bias . 15 Characteristics of the Relative Risk When Random Sampling . and Not . 19 Types of Bias . 20 Selection Bias . 21 Information Bias . 23 Conclusion . 24 Take-Home, Open-Book Quiz (Student Version) . 25 Take-Home, Open-Book Quiz (Teacher’s Answer Key) . 27 In-Class Exercise (Student Version) . 30 In-Class Exercise (Teacher’s Answer Key) . 32 Bias in Epidemiologic Research (Examination) (Student Version) . 33 Bias in Epidemiologic Research (Examination with Answers) (Teacher’s Answer Key) . 35 Copyright © 2004 by College Entrance Examination Board. All rights reserved. College Board, SAT and the acorn logo are registered trademarks of the College Entrance Examination Board. Other products and services may be trademarks of their respective owners. Visit College Board on the Web: www.collegeboard.com. Copyright © 2004. All rights reserved. 2 Observational Studies and Bias in Epidemiology Lesson Plan TITLE: Observational Studies and Bias in Epidemiology SUBJECT AREA: Biology, mathematics, statistics, environmental and health sciences GOAL: To identify and appreciate the effects of bias in epidemiologic research OBJECTIVES: 1. Introduce students to the principles and methods for interpreting the results of epidemio- logic research and bias 2.
    [Show full text]
  • Statistical Tips for Interpreting Scientific Claims
    Research Skills Seminar Series 2019 CAHS Research Education Program Statistical Tips for Interpreting Scientific Claims Mark Jones Statistician, Telethon Kids Institute 18 October 2019 Research Skills Seminar Series | CAHS Research Education Program Department of Child Health Research | Child and Adolescent Health Service ResearchEducationProgram.org © CAHS Research Education Program, Department of Child Health Research, Child and Adolescent Health Service, WA 2019 Copyright to this material produced by the CAHS Research Education Program, Department of Child Health Research, Child and Adolescent Health Service, Western Australia, under the provisions of the Copyright Act 1968 (C’wth Australia). Apart from any fair dealing for personal, academic, research or non-commercial use, no part may be reproduced without written permission. The Department of Child Health Research is under no obligation to grant this permission. Please acknowledge the CAHS Research Education Program, Department of Child Health Research, Child and Adolescent Health Service when reproducing or quoting material from this source. Statistical Tips for Interpreting Scientific Claims CONTENTS: 1 PRESENTATION ............................................................................................................................... 1 2 ARTICLE: TWENTY TIPS FOR INTERPRETING SCIENTIFIC CLAIMS, SUTHERLAND, SPIEGELHALTER & BURGMAN, 2013 .................................................................................................................................. 15 3
    [Show full text]
  • Correcting Sampling Bias in Non-Market Valuation with Kernel Mean Matching
    CORRECTING SAMPLING BIAS IN NON-MARKET VALUATION WITH KERNEL MEAN MATCHING Rui Zhang Department of Agricultural and Applied Economics University of Georgia [email protected] Selected Paper prepared for presentation at the 2017 Agricultural & Applied Economics Association Annual Meeting, Chicago, Illinois, July 30 - August 1 Copyright 2017 by Rui Zhang. All rights reserved. Readers may make verbatim copies of this document for non-commercial purposes by any means, provided that this copyright notice appears on all such copies. Abstract Non-response is common in surveys used in non-market valuation studies and can bias the parameter estimates and mean willingness to pay (WTP) estimates. One approach to correct this bias is to reweight the sample so that the distribution of the characteristic variables of the sample can match that of the population. We use a machine learning algorism Kernel Mean Matching (KMM) to produce resampling weights in a non-parametric manner. We test KMM’s performance through Monte Carlo simulations under multiple scenarios and show that KMM can effectively correct mean WTP estimates, especially when the sample size is small and sampling process depends on covariates. We also confirm KMM’s robustness to skewed bid design and model misspecification. Key Words: contingent valuation, Kernel Mean Matching, non-response, bias correction, willingness to pay 2 1. Introduction Nonrandom sampling can bias the contingent valuation estimates in two ways. Firstly, when the sample selection process depends on the covariate, the WTP estimates are biased due to the divergence between the covariate distributions of the sample and the population, even the parameter estimates are consistent; this is usually called non-response bias.
    [Show full text]