The paper introduces, compares and contrasts formal models of source reliability proposed in the ... more The paper introduces, compares and contrasts formal models of source reliability proposed in the epistemology literature, in particular the prominent models of Bovens and Hartmann (2003) and Olsson (2011). All are Bayesian models seeking to provide normative guidance, yet they differ subtly in assumptions and resulting behavior. Models are evaluated both on conceptual grounds and through simulations, and the relationship between models is clarified. The simulations both show surprising similarities and highlight relevant differences between these models. Most importantly, however, our evaluations reveal that important normative concerns arguably remain unresolved. The philosophical implications of this for testimony are discussed.
Proceedings of the Fourty-First Annual Conference of the Cognitive Science Society, 2019
We start with the distinction of outcome- and belief-based Bayesian models of the sequential upda... more We start with the distinction of outcome- and belief-based Bayesian models of the sequential update of agents’ beliefs and subjective reliability of sources (trust). We then focus on discussing the influential Bayesian model of belief-based trust update by Eric Olsson, which models dichotomic events and explicitly represents anti-reliability. After sketching some disastrous recent results for this perhaps most promising model of belief update, we show new simulation results for the temporal dynamics of learning belief with and without trust update and with and without communication. The results seem to shed at least a somewhat more positive light on the communicating-and-trust-updating agents. This may be a light at the end of the tunnel of belief-based models of trust updating, but the interpretation of the clear findings is much less clear.
Journal of Experimental Psychology: Applied, Jan 23, 2019
In social-dilemma situations (e.g., public-good games), people may pursue their local self-intere... more In social-dilemma situations (e.g., public-good games), people may pursue their local self-interests, thereby lowering the overall payoff of their group and, paradoxically, even their individual payoffs as a result. Likewise, in inner-individual dilemmas, even without conflict of interest between persons, people may pursue local goals at the expense of overall utility. Our experiments investigate such dissociations of individual- and group-level effects in the context of personnel evaluation and selection. Participants were given the role of human resource managers selecting workers to optimize the overall payoff for the company. We investigated contexts where the individually best/worst employees systematically caused the worst/best group performance. When workers in a team could substantially increase or decrease coworkers performance, most participants (albeit not all) tended to focus solely on individual performance without considering their overall contribution when instructed to maximize group performance. This undue focus on individual information meant that employees who enhanced team performance the most often received the most negative evaluations. This may result in a ‘tragedy of personnel evaluation’ relevant to maladaptive incentive structures (personnel evaluation), job offers (personnel selection), and a substantially negative impact on organizational effectiveness. At the same time, the results suggest ways this problem may be overcome.
Proceedings of the Fourty-First Annual Conference of the Cognitive Science Society, 2019
Social dilemmas conceptually suggest distinguishing direct individual and group-level effects (al... more Social dilemmas conceptually suggest distinguishing direct individual and group-level effects (also involving indirect effects on others). Furthermore, the success of organizations appears to rely on identifying not only individual excellence but positive impact on others as well. In ‘Two-Level Personnel Evaluation Tasks’ (T-PETs) participants as human resource managers evaluate employees when individual and group contributions are dissociated. Von Sydow, Braus, & Hahn (2018) have suggested a potential ‘Tragedy of Personnel Evaluation’: A group-serving employee with the smallest individual contribution but by far the greatest po-sitive effect on the group’s overall earnings was often rated the most negatively. Here we investigate, in two experiments with conflicting information, whether emphasizing the group can avert the ‘tragic’ outcome. Our results suggest that the tragedy is not as complete as suggested, and that contextual information can mitigate the tragedy. Nonetheless, the results also corroborate the stability of underestimating the impact of team players
With the advent of social media, the last decade has seen profound changes to the way people rece... more With the advent of social media, the last decade has seen profound changes to the way people receive information. This has fueled debate about the ways (if any) changes to the nature of our information networks might be affecting voters' beliefs about the world, voting results, and, ultimately, democracy. At the same time, much discussion in the public arena in recent years has concerned the notion that ill-informed voters have been voting against their own self-interest. The research reported here brings these two strands together: simulations involving agent-based models, interpreted through the formal framework of Condorcet's (1785) Jury Theorem, demonstrate how changes to information networks may make voter error more likely even though individual competence has largely remained unchanged.
In social-dilemma situations (e.g., public-good games), people may pursue their local self-intere... more In social-dilemma situations (e.g., public-good games), people may pursue their local self-interests, thereby lowering the overall payoff of their group and, paradoxically, even their individual payoffs as a result. Likewise, in inner-individual dilemmas, even without conflict of interest between persons, people may pursue local goals at the expense of overall utility. Our experiments investigate such dissociations of individual and group-level effects in the context of personnel evaluation and selection. Participants were given the role of human resource managers selecting workers to optimize the overall payoff for the company. We investigated contexts where the individually best/worst ‘employees’ systematically caused the worst/best group performance. When workers in a team could substantially increase or decrease co-workers’ performance, most participants (albeit not all) tended to focus solely on individual performance without considering their overall contribution even when instructed to maximize group performance. This undue focus on individual information meant that employees who enhanced team performance the most often received the most negative evaluations. This may result in a ‘tragedy of personnel evaluation’ relevant to maladaptive incentive structures (personnel evaluation), job offers (personnel selection), and a substantially negative impact on organizational effectiveness. At the same time, the results suggest ways this problem may be overcome.
This paper examines the basic question of how we can come to form accurate beliefs about the worl... more This paper examines the basic question of how we can come to form accurate beliefs about the world when we do not fully know how good or bad our evidence is. Here we show, using simulations with otherwise optimal agents, the cost of misjudging the quality of our evidence, and compare different strategies for correctly estimating that quality, such as outcome, and expectation-based updating. We identify conditions under which misjudgment of evidence quality can nevertheless lead to accurate beliefs, as well as those conditions where no strategy will help. These results indicate both where people will nevertheless succeed and where they will fail when information quality is degraded. Many things about the world we cannot observe directly. These range from the mundane ('has the dog stolen the paper?', 'are these symptoms caused by flu?') to the complex, overarching theories of science (evolution, quantum mechanics and so on). The 'knowledge' we believe to have about these things rests crucially on inference. Even what we consider to be 'direct observation' of the world (such as the relative depth of objects in our visual field) is the result of constructive, inferential processes that create a 'model' of the world based on the perceptual evidence available [1]. Any such inference faces the fundamental question of the extent to which a given body of evidence licenses the conclusion we wish to draw. For certain contexts, this question has a clear, normative, answer in Bayesian belief updating, as we will outline below; but the accuracy of such inference will be limited by our understanding of the quality of the data: how well do they indicate the truth or falsity of our hypothesis? Crucially, in many real-world situations these data characteristics will not be known exactly or even not be known at all, as in the case of the stranger giving us directions, the witness in court, or the social media post about a politician's behaviour. What
Causal reasoning is crucial to people’s decision-making in probabilistic environments. It may re... more Causal reasoning is crucial to people’s decision-making in probabilistic environments. It may rely directly on data about covariation between variables (correspondence) or on inferences based on reasonable constraints if larger causal models are constructed based on local relations (coherence). For causal chains an often assumed constraint is transitivity. For probabilistic causal relations, mismatches between such transitive inferences and direct empirical evidence may lead to distortions of empirical evidence. Previous work has shown that people may use the generative local causal relations
A → B and B → C to infer a positive indirect relation between events A and C, despite data showing that these events are actually independent (von Sydow et al., 2009, 2010, 2016). Here we used an economic sequential learning scenario to investigate how transitive reasoning in intransitive situations with negatively related distal events may relate to betting behavior. In three experiments participants bet as if they were influenced by a transitivity assumption, even when the data strongly contradicted transitivity.
Probability judgments entail a conjunction fallacy (CF) if a conjunction is estimated to be more ... more Probability judgments entail a conjunction fallacy (CF) if a conjunction is estimated to be more probable than one of its conjuncts. In the context of predication of alternative logical hypothesis, Bayesian logic (BL) provides a formalization of pattern probabilities that renders a class of pattern-based CFs rational. BL predicts a complete system of other logical inclusion fallacies (IFs). A first test of this prediction is investigated here, using transparent tasks with clear set-inclusions, varying in observed frequencies only. Experiment 1 uses data where BL makes dominant predictions; Experiment 2’s predictions were less clear, and we additionally investigated judgments about second-most probable hypotheses. The results corroborated a pattern-probability account and cannot be easily explained by other theories of CFs (e.g., inverse probability, confirmation). IFs were not limited to conjunctions, but rather occurred systematically for several logical connectives. Thus pattern-based probability judgements about logical relations may constitute a basic class of intuitive but potentially rational probability judgements.
A probabilistic causal chain A -> B -> C may intuitively appear to be transitive: If A probabilis... more A probabilistic causal chain A -> B -> C may intuitively appear to be transitive: If A probabilistically causes B, and B probabilistically causes C, A probabilistically causes
C. However, probabilistic causal relations are only transitive if the so-called Markov condition holds. In two experiments, we examined how people make probabilistic judgments about indirect relationships A -> C in causal chains A -> B -> C that violate the Markov condition. We hypothesized that participants would make transitive inferences in accordance with the Markov condition although they were presented with counterevidence showing intransitive data. For instance, participants were successively presented with data entailing positive dependencies
A -> B and B -> C. At the same time, the data entailed that A and C were statistically independent. The results of two experiments show that transitive reasoning via a mediating event B influenced and distorted the induction of the indirect relation between A and C. Participants’ judgments were affected by an interaction of transitive, causal-model-based inferences and the observed data. Our findings support the idea that people tend to chain individual causal relations into mental causal chains that obey the Markov condition and thus allow for transitive reasoning, even if the observed data entail that such inferences are not warranted.
In: M. W. Eysenck & D. Groome. Cognitive Psychology: Revisiting the Classical Studies. Los Angeles, London: Sage., 2015
It is no exaggeration to say that today’s psychology would not be what it is without Daniel Kahn... more It is no exaggeration to say that today’s psychology would not be what it is without Daniel Kahneman’s and Amos Tversky’s seminal work on heuristics and biases, as summarised in a Science article (Tversky & Kahneman, 1974) that was cited over 7,000 times – an unbelievable rate for a psychology article. A few years before this work spread like wildfire. The rationalist metaphor of a computer-like human mem-ory and the man-as-scientist analogy conveyed in theories of consistency (Abelson, 1968) and attribution (Jones, Kanouse, Kelley, Nisbett, Valins, & Weiner, 1987) had brought about the so-called cognitive revolution (Dember, 1974). However, this naïvely optimistic view on the human mind turned rapidly into deflating pessimism when statistical tools began to dominate research and theorising came under the dominating influence of Kahneman and Tversky’s research programme.
An inductive, pattern-sensitive Bayesian logic (BL) is proposed as a normative and descriptive mo... more An inductive, pattern-sensitive Bayesian logic (BL) is proposed as a normative and descriptive model for probability judgments about hypotheses involving probabilistic logical connectives. The model explains a specific class of frequency-based conjunction fallacies (CFs). It is suggested that the pattern probabilities calculated by BL may serve as a criterion of noisy-logical predication, resolving some paradoxes of predication. The model is developed for frequency information in 2 × 2 contingency tables. According to standard probability theory, a violation of the conjunction rule, P(A) ≥ P(A ∧ B) (e.g., P(ravens are black) ≥ P(ravens are black AND they can fly)), is always a fallacy. A frequentist interpretation of probability has exculpated participants from committing CFs when one is concerned with single events. Here a pattern-based Bayesian interpretation of probabilities of (noisy) dyadic logical predications is elaborated, predicting frequency-based but rational 'CFs'. BL formalizes the probabilities of logical patterns, integrating over noise levels. BL, for instance, predicts double CFs, differential sample-size effects, and pattern sensitivity. Three experiments provide a first corroboration that BL is also an adequate empirical model to predict logical probability judgments based on 2 × 2 contingency tables. BL may shed light on the more general rationality debate.
Proceedings of the Thirthy-Eighth Annual Conference of the Cognitive Science Society, 2016
In social-dilemma situations (public-good games) people may pursue their local, egoistic interest... more In social-dilemma situations (public-good games) people may pursue their local, egoistic interests and thereby lower the global, overall payoff of their group and, paradoxically, even their own resulting payoff. One may also speak of intra-individual dilemmas, where people pursue local goals at the expense of their overall utility. Our current experiments transfer this idea to the context of personnel evaluation and personnel selection. In our experiments, participants were put in the position of a Human Resources manager, who should for instance select workers who optimize the overall payoff of the company, rather than those who optimize only their specific payoffs. The results of the experiments, however, suggest that most, albeit not all, participants tended to focus on directly comparing individuals without considering the overall contribution to a group. Thus employees with the best overall effects for a company or organization may be evaluated the most negatively. This possible 'tragedy of personnel evaluation' may be linked to maladaptive incentive structures (personnel evaluation), advancement of employees (personnel promotion) and job offers (personnel selection), and may have a substantial negative impact on the effectiveness of companies or organizations.
Proceedings of the Thirty-Ninth Annual Conference of the Cognitive Science Society, 2017
In the Wason-Selection Task debate it has been suggested that people may be able to detect cheate... more In the Wason-Selection Task debate it has been suggested that people may be able to detect cheaters but not co-operators or altruists. This position has been challenged. Here we focus on a scenario that is more ecologically valid with regard to different strategies for detecting workers who negatively interact with others (here 'egoists') and positive interactors (here 'altruist'). The results on altruist detection in two-level personnel evaluation tasks (T-PETs), with information on individual and team performance, suggested a disregard of the team performance and a resulting " Tragedy of Personnel Evaluation ". Experiment 1 transfers the idea of altruist detection in a personnel evaluation and personnel selection task (von Sydow & Braus, 2016) to egoist detection and explores whether there are analogous problems for egoist detection. Experiment 2 explores egoist and altruist detection in more realistic settings where individual and group-selection may affect our sampling of the interactor.
Proceedings of the Thirty-Ninth Annual Conference of the Cognitive Science Society, 2017
Human beings are essentially – by nature or second nature-members of groups. They contribute to t... more Human beings are essentially – by nature or second nature-members of groups. They contribute to these groups not just as isolated individuals but also through their interaction with others. Consequently, personnel evaluation in companies and organizations requires assessing not only evaluating individual performance but also the overall direct and indirect effect one has on a team. Others' work may be improved or hampered by the presence of a particular employee. We investigate Two-level Personnel-Evaluation Tasks (T-PETs) with information on individual and group earnings, where an individual focus may lead to evaluate the overall best employee as being the worst. We have previously found a Tragedy of Personnel Evaluation where focus on direct individual impact did have such systematic effect. In two experiments, one on team size, the other on kinds of information provided, we explore the boundary conditions of this effect and suggest how it may be overcome.
Proceedings of the Thirty-Ninth Annual Conference of the Cognitive Science Society, 2017
Conjunction fallacies (CF) have not only been a major obstacle in justifying the rationality of a... more Conjunction fallacies (CF) have not only been a major obstacle in justifying the rationality of a Bayesian theory of belief update; they have also inspired a variety of theories on probability judgment and logical predication. Here we provide an overview of Bayesian logic (BL) as rational formulation of a pattern-based class of conjunction fallacies. BL is described here as a generalization of Bayesian Occam's razor. BL captures the idea that probabilities are sometimes used not extensionally but intensionally, determining the probabilistic adequacy of ideal logical patterns. It is emphasized that BL is a class of models that depend on representations and the meanings of logical connectives. We discuss open questions and limits of BL. We also briefly discuss whether other theories of the CF may be good supplementary theories of CFs (and predication) as well, if linked to functional explanations.
This work proposes and elaborates a philosophy of nature that, although influenced
by Darwinism... more This work proposes and elaborates a philosophy of nature that, although influenced
by Darwinism, aims ultimately to transcend Darwinism. My particular focus is on two
purified versions of Darwinism: gene-Darwinism and process-Darwinism. The essential claims of these two approaches are first explicated and then subjected to criticism. This elaborated critique is not exogeneous to Darwinism, proposing another philosophy of nature from the outset; instead an immanent critique is developed, starting from within the investigated Darwinian paradigms. Focussing on internal inconsistencies of these
paradigms, reveals tendencies that will lead us beyond Darwinism.But not only theories can transcend themselves, the central claim of this work is that Nature, due to inner or outer necessities, continually transcends itself, not only in its products but in its evolutionary mechanisms. As theories are moulded not only by external forces, but by inherent tendencies as well (where the rules of change may sometimes depend on the theory itself), also evolution may depend on evolved evolutionary mechanisms.
In E. Voigts, B. Schaff &M. Pietrzak-Franger (Eds.). Reflecting on Darwin. Farnham, London: Ashgate., 2014
Conclusion: First, it was shown that tautological interpretations of 'survival of the fittest ', ... more Conclusion: First, it was shown that tautological interpretations of 'survival of the fittest ', based on defining the explanandum by the explanans, are surprisingly stable against several modifications of the meaning of the term 'fitness'. Simultaneous shifts on the sides of both the explanans and the explanandum were shown to be possible. If natural selection in biology is defined in this circular way it can never be refuted and is at best a metaphysical principle. Second, it was argued that one may nonetheless provide testable definitions of natural selection, based explicitly on using the concepts of blind variation (blindness) and environmental selection (external ism) in its definition. Although taking the tautology problem seriously, and pointing out that even the criteria cited are not trivial to operationalize, it was defended that 'survival of the fittest' may be formulated in a testable way. The testable formulations, however, may actually lead to a falsification of natural selection or to restricting its domain of application. Finally, it was argued that in reinforcement learning, a Darwinian process analogous to natural selection, the problem of tautology can be discussed in an analogous way as well. Again much care is needed to disentangle tautological from testable aspects. Only then can one obtain a truly empirical theory that may indeed turn out to be false or at least incomplete. Alternatively, one may of course treat these theories as non-empirical etaphysical frameworks only, generating empirical hypotheses and contributing to a larger Darwinian metaphysics without being testable. Nevertheless, an implicit shifting between a testable and an untestable interpretation can be an illicit tactic to immunize natural selection or reinforcement learning while conveying the impression that one is concerned with testable hypotheses.
The paper introduces, compares and contrasts formal models of source reliability proposed in the ... more The paper introduces, compares and contrasts formal models of source reliability proposed in the epistemology literature, in particular the prominent models of Bovens and Hartmann (2003) and Olsson (2011). All are Bayesian models seeking to provide normative guidance, yet they differ subtly in assumptions and resulting behavior. Models are evaluated both on conceptual grounds and through simulations, and the relationship between models is clarified. The simulations both show surprising similarities and highlight relevant differences between these models. Most importantly, however, our evaluations reveal that important normative concerns arguably remain unresolved. The philosophical implications of this for testimony are discussed.
Proceedings of the Fourty-First Annual Conference of the Cognitive Science Society, 2019
We start with the distinction of outcome- and belief-based Bayesian models of the sequential upda... more We start with the distinction of outcome- and belief-based Bayesian models of the sequential update of agents’ beliefs and subjective reliability of sources (trust). We then focus on discussing the influential Bayesian model of belief-based trust update by Eric Olsson, which models dichotomic events and explicitly represents anti-reliability. After sketching some disastrous recent results for this perhaps most promising model of belief update, we show new simulation results for the temporal dynamics of learning belief with and without trust update and with and without communication. The results seem to shed at least a somewhat more positive light on the communicating-and-trust-updating agents. This may be a light at the end of the tunnel of belief-based models of trust updating, but the interpretation of the clear findings is much less clear.
Journal of Experimental Psychology: Applied, Jan 23, 2019
In social-dilemma situations (e.g., public-good games), people may pursue their local self-intere... more In social-dilemma situations (e.g., public-good games), people may pursue their local self-interests, thereby lowering the overall payoff of their group and, paradoxically, even their individual payoffs as a result. Likewise, in inner-individual dilemmas, even without conflict of interest between persons, people may pursue local goals at the expense of overall utility. Our experiments investigate such dissociations of individual- and group-level effects in the context of personnel evaluation and selection. Participants were given the role of human resource managers selecting workers to optimize the overall payoff for the company. We investigated contexts where the individually best/worst employees systematically caused the worst/best group performance. When workers in a team could substantially increase or decrease coworkers performance, most participants (albeit not all) tended to focus solely on individual performance without considering their overall contribution when instructed to maximize group performance. This undue focus on individual information meant that employees who enhanced team performance the most often received the most negative evaluations. This may result in a ‘tragedy of personnel evaluation’ relevant to maladaptive incentive structures (personnel evaluation), job offers (personnel selection), and a substantially negative impact on organizational effectiveness. At the same time, the results suggest ways this problem may be overcome.
Proceedings of the Fourty-First Annual Conference of the Cognitive Science Society, 2019
Social dilemmas conceptually suggest distinguishing direct individual and group-level effects (al... more Social dilemmas conceptually suggest distinguishing direct individual and group-level effects (also involving indirect effects on others). Furthermore, the success of organizations appears to rely on identifying not only individual excellence but positive impact on others as well. In ‘Two-Level Personnel Evaluation Tasks’ (T-PETs) participants as human resource managers evaluate employees when individual and group contributions are dissociated. Von Sydow, Braus, & Hahn (2018) have suggested a potential ‘Tragedy of Personnel Evaluation’: A group-serving employee with the smallest individual contribution but by far the greatest po-sitive effect on the group’s overall earnings was often rated the most negatively. Here we investigate, in two experiments with conflicting information, whether emphasizing the group can avert the ‘tragic’ outcome. Our results suggest that the tragedy is not as complete as suggested, and that contextual information can mitigate the tragedy. Nonetheless, the results also corroborate the stability of underestimating the impact of team players
With the advent of social media, the last decade has seen profound changes to the way people rece... more With the advent of social media, the last decade has seen profound changes to the way people receive information. This has fueled debate about the ways (if any) changes to the nature of our information networks might be affecting voters' beliefs about the world, voting results, and, ultimately, democracy. At the same time, much discussion in the public arena in recent years has concerned the notion that ill-informed voters have been voting against their own self-interest. The research reported here brings these two strands together: simulations involving agent-based models, interpreted through the formal framework of Condorcet's (1785) Jury Theorem, demonstrate how changes to information networks may make voter error more likely even though individual competence has largely remained unchanged.
In social-dilemma situations (e.g., public-good games), people may pursue their local self-intere... more In social-dilemma situations (e.g., public-good games), people may pursue their local self-interests, thereby lowering the overall payoff of their group and, paradoxically, even their individual payoffs as a result. Likewise, in inner-individual dilemmas, even without conflict of interest between persons, people may pursue local goals at the expense of overall utility. Our experiments investigate such dissociations of individual and group-level effects in the context of personnel evaluation and selection. Participants were given the role of human resource managers selecting workers to optimize the overall payoff for the company. We investigated contexts where the individually best/worst ‘employees’ systematically caused the worst/best group performance. When workers in a team could substantially increase or decrease co-workers’ performance, most participants (albeit not all) tended to focus solely on individual performance without considering their overall contribution even when instructed to maximize group performance. This undue focus on individual information meant that employees who enhanced team performance the most often received the most negative evaluations. This may result in a ‘tragedy of personnel evaluation’ relevant to maladaptive incentive structures (personnel evaluation), job offers (personnel selection), and a substantially negative impact on organizational effectiveness. At the same time, the results suggest ways this problem may be overcome.
This paper examines the basic question of how we can come to form accurate beliefs about the worl... more This paper examines the basic question of how we can come to form accurate beliefs about the world when we do not fully know how good or bad our evidence is. Here we show, using simulations with otherwise optimal agents, the cost of misjudging the quality of our evidence, and compare different strategies for correctly estimating that quality, such as outcome, and expectation-based updating. We identify conditions under which misjudgment of evidence quality can nevertheless lead to accurate beliefs, as well as those conditions where no strategy will help. These results indicate both where people will nevertheless succeed and where they will fail when information quality is degraded. Many things about the world we cannot observe directly. These range from the mundane ('has the dog stolen the paper?', 'are these symptoms caused by flu?') to the complex, overarching theories of science (evolution, quantum mechanics and so on). The 'knowledge' we believe to have about these things rests crucially on inference. Even what we consider to be 'direct observation' of the world (such as the relative depth of objects in our visual field) is the result of constructive, inferential processes that create a 'model' of the world based on the perceptual evidence available [1]. Any such inference faces the fundamental question of the extent to which a given body of evidence licenses the conclusion we wish to draw. For certain contexts, this question has a clear, normative, answer in Bayesian belief updating, as we will outline below; but the accuracy of such inference will be limited by our understanding of the quality of the data: how well do they indicate the truth or falsity of our hypothesis? Crucially, in many real-world situations these data characteristics will not be known exactly or even not be known at all, as in the case of the stranger giving us directions, the witness in court, or the social media post about a politician's behaviour. What
Causal reasoning is crucial to people’s decision-making in probabilistic environments. It may re... more Causal reasoning is crucial to people’s decision-making in probabilistic environments. It may rely directly on data about covariation between variables (correspondence) or on inferences based on reasonable constraints if larger causal models are constructed based on local relations (coherence). For causal chains an often assumed constraint is transitivity. For probabilistic causal relations, mismatches between such transitive inferences and direct empirical evidence may lead to distortions of empirical evidence. Previous work has shown that people may use the generative local causal relations
A → B and B → C to infer a positive indirect relation between events A and C, despite data showing that these events are actually independent (von Sydow et al., 2009, 2010, 2016). Here we used an economic sequential learning scenario to investigate how transitive reasoning in intransitive situations with negatively related distal events may relate to betting behavior. In three experiments participants bet as if they were influenced by a transitivity assumption, even when the data strongly contradicted transitivity.
Probability judgments entail a conjunction fallacy (CF) if a conjunction is estimated to be more ... more Probability judgments entail a conjunction fallacy (CF) if a conjunction is estimated to be more probable than one of its conjuncts. In the context of predication of alternative logical hypothesis, Bayesian logic (BL) provides a formalization of pattern probabilities that renders a class of pattern-based CFs rational. BL predicts a complete system of other logical inclusion fallacies (IFs). A first test of this prediction is investigated here, using transparent tasks with clear set-inclusions, varying in observed frequencies only. Experiment 1 uses data where BL makes dominant predictions; Experiment 2’s predictions were less clear, and we additionally investigated judgments about second-most probable hypotheses. The results corroborated a pattern-probability account and cannot be easily explained by other theories of CFs (e.g., inverse probability, confirmation). IFs were not limited to conjunctions, but rather occurred systematically for several logical connectives. Thus pattern-based probability judgements about logical relations may constitute a basic class of intuitive but potentially rational probability judgements.
A probabilistic causal chain A -> B -> C may intuitively appear to be transitive: If A probabilis... more A probabilistic causal chain A -> B -> C may intuitively appear to be transitive: If A probabilistically causes B, and B probabilistically causes C, A probabilistically causes
C. However, probabilistic causal relations are only transitive if the so-called Markov condition holds. In two experiments, we examined how people make probabilistic judgments about indirect relationships A -> C in causal chains A -> B -> C that violate the Markov condition. We hypothesized that participants would make transitive inferences in accordance with the Markov condition although they were presented with counterevidence showing intransitive data. For instance, participants were successively presented with data entailing positive dependencies
A -> B and B -> C. At the same time, the data entailed that A and C were statistically independent. The results of two experiments show that transitive reasoning via a mediating event B influenced and distorted the induction of the indirect relation between A and C. Participants’ judgments were affected by an interaction of transitive, causal-model-based inferences and the observed data. Our findings support the idea that people tend to chain individual causal relations into mental causal chains that obey the Markov condition and thus allow for transitive reasoning, even if the observed data entail that such inferences are not warranted.
In: M. W. Eysenck & D. Groome. Cognitive Psychology: Revisiting the Classical Studies. Los Angeles, London: Sage., 2015
It is no exaggeration to say that today’s psychology would not be what it is without Daniel Kahn... more It is no exaggeration to say that today’s psychology would not be what it is without Daniel Kahneman’s and Amos Tversky’s seminal work on heuristics and biases, as summarised in a Science article (Tversky & Kahneman, 1974) that was cited over 7,000 times – an unbelievable rate for a psychology article. A few years before this work spread like wildfire. The rationalist metaphor of a computer-like human mem-ory and the man-as-scientist analogy conveyed in theories of consistency (Abelson, 1968) and attribution (Jones, Kanouse, Kelley, Nisbett, Valins, & Weiner, 1987) had brought about the so-called cognitive revolution (Dember, 1974). However, this naïvely optimistic view on the human mind turned rapidly into deflating pessimism when statistical tools began to dominate research and theorising came under the dominating influence of Kahneman and Tversky’s research programme.
An inductive, pattern-sensitive Bayesian logic (BL) is proposed as a normative and descriptive mo... more An inductive, pattern-sensitive Bayesian logic (BL) is proposed as a normative and descriptive model for probability judgments about hypotheses involving probabilistic logical connectives. The model explains a specific class of frequency-based conjunction fallacies (CFs). It is suggested that the pattern probabilities calculated by BL may serve as a criterion of noisy-logical predication, resolving some paradoxes of predication. The model is developed for frequency information in 2 × 2 contingency tables. According to standard probability theory, a violation of the conjunction rule, P(A) ≥ P(A ∧ B) (e.g., P(ravens are black) ≥ P(ravens are black AND they can fly)), is always a fallacy. A frequentist interpretation of probability has exculpated participants from committing CFs when one is concerned with single events. Here a pattern-based Bayesian interpretation of probabilities of (noisy) dyadic logical predications is elaborated, predicting frequency-based but rational 'CFs'. BL formalizes the probabilities of logical patterns, integrating over noise levels. BL, for instance, predicts double CFs, differential sample-size effects, and pattern sensitivity. Three experiments provide a first corroboration that BL is also an adequate empirical model to predict logical probability judgments based on 2 × 2 contingency tables. BL may shed light on the more general rationality debate.
Proceedings of the Thirthy-Eighth Annual Conference of the Cognitive Science Society, 2016
In social-dilemma situations (public-good games) people may pursue their local, egoistic interest... more In social-dilemma situations (public-good games) people may pursue their local, egoistic interests and thereby lower the global, overall payoff of their group and, paradoxically, even their own resulting payoff. One may also speak of intra-individual dilemmas, where people pursue local goals at the expense of their overall utility. Our current experiments transfer this idea to the context of personnel evaluation and personnel selection. In our experiments, participants were put in the position of a Human Resources manager, who should for instance select workers who optimize the overall payoff of the company, rather than those who optimize only their specific payoffs. The results of the experiments, however, suggest that most, albeit not all, participants tended to focus on directly comparing individuals without considering the overall contribution to a group. Thus employees with the best overall effects for a company or organization may be evaluated the most negatively. This possible 'tragedy of personnel evaluation' may be linked to maladaptive incentive structures (personnel evaluation), advancement of employees (personnel promotion) and job offers (personnel selection), and may have a substantial negative impact on the effectiveness of companies or organizations.
Proceedings of the Thirty-Ninth Annual Conference of the Cognitive Science Society, 2017
In the Wason-Selection Task debate it has been suggested that people may be able to detect cheate... more In the Wason-Selection Task debate it has been suggested that people may be able to detect cheaters but not co-operators or altruists. This position has been challenged. Here we focus on a scenario that is more ecologically valid with regard to different strategies for detecting workers who negatively interact with others (here 'egoists') and positive interactors (here 'altruist'). The results on altruist detection in two-level personnel evaluation tasks (T-PETs), with information on individual and team performance, suggested a disregard of the team performance and a resulting " Tragedy of Personnel Evaluation ". Experiment 1 transfers the idea of altruist detection in a personnel evaluation and personnel selection task (von Sydow & Braus, 2016) to egoist detection and explores whether there are analogous problems for egoist detection. Experiment 2 explores egoist and altruist detection in more realistic settings where individual and group-selection may affect our sampling of the interactor.
Proceedings of the Thirty-Ninth Annual Conference of the Cognitive Science Society, 2017
Human beings are essentially – by nature or second nature-members of groups. They contribute to t... more Human beings are essentially – by nature or second nature-members of groups. They contribute to these groups not just as isolated individuals but also through their interaction with others. Consequently, personnel evaluation in companies and organizations requires assessing not only evaluating individual performance but also the overall direct and indirect effect one has on a team. Others' work may be improved or hampered by the presence of a particular employee. We investigate Two-level Personnel-Evaluation Tasks (T-PETs) with information on individual and group earnings, where an individual focus may lead to evaluate the overall best employee as being the worst. We have previously found a Tragedy of Personnel Evaluation where focus on direct individual impact did have such systematic effect. In two experiments, one on team size, the other on kinds of information provided, we explore the boundary conditions of this effect and suggest how it may be overcome.
Proceedings of the Thirty-Ninth Annual Conference of the Cognitive Science Society, 2017
Conjunction fallacies (CF) have not only been a major obstacle in justifying the rationality of a... more Conjunction fallacies (CF) have not only been a major obstacle in justifying the rationality of a Bayesian theory of belief update; they have also inspired a variety of theories on probability judgment and logical predication. Here we provide an overview of Bayesian logic (BL) as rational formulation of a pattern-based class of conjunction fallacies. BL is described here as a generalization of Bayesian Occam's razor. BL captures the idea that probabilities are sometimes used not extensionally but intensionally, determining the probabilistic adequacy of ideal logical patterns. It is emphasized that BL is a class of models that depend on representations and the meanings of logical connectives. We discuss open questions and limits of BL. We also briefly discuss whether other theories of the CF may be good supplementary theories of CFs (and predication) as well, if linked to functional explanations.
This work proposes and elaborates a philosophy of nature that, although influenced
by Darwinism... more This work proposes and elaborates a philosophy of nature that, although influenced
by Darwinism, aims ultimately to transcend Darwinism. My particular focus is on two
purified versions of Darwinism: gene-Darwinism and process-Darwinism. The essential claims of these two approaches are first explicated and then subjected to criticism. This elaborated critique is not exogeneous to Darwinism, proposing another philosophy of nature from the outset; instead an immanent critique is developed, starting from within the investigated Darwinian paradigms. Focussing on internal inconsistencies of these
paradigms, reveals tendencies that will lead us beyond Darwinism.But not only theories can transcend themselves, the central claim of this work is that Nature, due to inner or outer necessities, continually transcends itself, not only in its products but in its evolutionary mechanisms. As theories are moulded not only by external forces, but by inherent tendencies as well (where the rules of change may sometimes depend on the theory itself), also evolution may depend on evolved evolutionary mechanisms.
In E. Voigts, B. Schaff &M. Pietrzak-Franger (Eds.). Reflecting on Darwin. Farnham, London: Ashgate., 2014
Conclusion: First, it was shown that tautological interpretations of 'survival of the fittest ', ... more Conclusion: First, it was shown that tautological interpretations of 'survival of the fittest ', based on defining the explanandum by the explanans, are surprisingly stable against several modifications of the meaning of the term 'fitness'. Simultaneous shifts on the sides of both the explanans and the explanandum were shown to be possible. If natural selection in biology is defined in this circular way it can never be refuted and is at best a metaphysical principle. Second, it was argued that one may nonetheless provide testable definitions of natural selection, based explicitly on using the concepts of blind variation (blindness) and environmental selection (external ism) in its definition. Although taking the tautology problem seriously, and pointing out that even the criteria cited are not trivial to operationalize, it was defended that 'survival of the fittest' may be formulated in a testable way. The testable formulations, however, may actually lead to a falsification of natural selection or to restricting its domain of application. Finally, it was argued that in reinforcement learning, a Darwinian process analogous to natural selection, the problem of tautology can be discussed in an analogous way as well. Again much care is needed to disentangle tautological from testable aspects. Only then can one obtain a truly empirical theory that may indeed turn out to be false or at least incomplete. Alternatively, one may of course treat these theories as non-empirical etaphysical frameworks only, generating empirical hypotheses and contributing to a larger Darwinian metaphysics without being testable. Nevertheless, an implicit shifting between a testable and an untestable interpretation can be an illicit tactic to immunize natural selection or reinforcement learning while conveying the impression that one is concerned with testable hypotheses.
Uploads
Papers by Momme von Sydow
A → B and B → C to infer a positive indirect relation between events A and C, despite data showing that these events are actually independent (von Sydow et al., 2009, 2010, 2016). Here we used an economic sequential learning scenario to investigate how transitive reasoning in intransitive situations with negatively related distal events may relate to betting behavior. In three experiments participants bet as if they were influenced by a transitivity assumption, even when the data strongly contradicted transitivity.
C. However, probabilistic causal relations are only transitive if the so-called Markov condition holds. In two experiments, we examined how people make probabilistic judgments about indirect relationships A -> C in causal chains A -> B -> C that violate the Markov condition. We hypothesized that participants would make transitive inferences in accordance with the Markov condition although they were presented with counterevidence showing intransitive data. For instance, participants were successively presented with data entailing positive dependencies
A -> B and B -> C. At the same time, the data entailed that A and C were statistically independent. The results of two experiments show that transitive reasoning via a mediating event B influenced and distorted the induction of the indirect relation between A and C. Participants’ judgments were affected by an interaction of transitive, causal-model-based inferences and the observed data. Our findings support the idea that people tend to chain individual causal relations into mental causal chains that obey the Markov condition and thus allow for transitive reasoning, even if the observed data entail that such inferences are not warranted.
by Darwinism, aims ultimately to transcend Darwinism. My particular focus is on two
purified versions of Darwinism: gene-Darwinism and process-Darwinism. The essential claims of these two approaches are first explicated and then subjected to criticism. This elaborated critique is not exogeneous to Darwinism, proposing another philosophy of nature from the outset; instead an immanent critique is developed, starting from within the investigated Darwinian paradigms. Focussing on internal inconsistencies of these
paradigms, reveals tendencies that will lead us beyond Darwinism.But not only theories can transcend themselves, the central claim of this work is that Nature, due to inner or outer necessities, continually transcends itself, not only in its products but in its evolutionary mechanisms. As theories are moulded not only by external forces, but by inherent tendencies as well (where the rules of change may sometimes depend on the theory itself), also evolution may depend on evolved evolutionary mechanisms.
A → B and B → C to infer a positive indirect relation between events A and C, despite data showing that these events are actually independent (von Sydow et al., 2009, 2010, 2016). Here we used an economic sequential learning scenario to investigate how transitive reasoning in intransitive situations with negatively related distal events may relate to betting behavior. In three experiments participants bet as if they were influenced by a transitivity assumption, even when the data strongly contradicted transitivity.
C. However, probabilistic causal relations are only transitive if the so-called Markov condition holds. In two experiments, we examined how people make probabilistic judgments about indirect relationships A -> C in causal chains A -> B -> C that violate the Markov condition. We hypothesized that participants would make transitive inferences in accordance with the Markov condition although they were presented with counterevidence showing intransitive data. For instance, participants were successively presented with data entailing positive dependencies
A -> B and B -> C. At the same time, the data entailed that A and C were statistically independent. The results of two experiments show that transitive reasoning via a mediating event B influenced and distorted the induction of the indirect relation between A and C. Participants’ judgments were affected by an interaction of transitive, causal-model-based inferences and the observed data. Our findings support the idea that people tend to chain individual causal relations into mental causal chains that obey the Markov condition and thus allow for transitive reasoning, even if the observed data entail that such inferences are not warranted.
by Darwinism, aims ultimately to transcend Darwinism. My particular focus is on two
purified versions of Darwinism: gene-Darwinism and process-Darwinism. The essential claims of these two approaches are first explicated and then subjected to criticism. This elaborated critique is not exogeneous to Darwinism, proposing another philosophy of nature from the outset; instead an immanent critique is developed, starting from within the investigated Darwinian paradigms. Focussing on internal inconsistencies of these
paradigms, reveals tendencies that will lead us beyond Darwinism.But not only theories can transcend themselves, the central claim of this work is that Nature, due to inner or outer necessities, continually transcends itself, not only in its products but in its evolutionary mechanisms. As theories are moulded not only by external forces, but by inherent tendencies as well (where the rules of change may sometimes depend on the theory itself), also evolution may depend on evolved evolutionary mechanisms.