Selection bias? Content Type Journal Article DOI 10.1007/s11016-010-9490-4 Authors Henry M. Cowles, Program in History of Science, Princeton University, Princeton, NJ 08544, USA Journal Metascience Online ISSN 1467-9981 Print ISSN 0815-0796.

Abstract:Suppose we want to know whether the ethics of persons with one characteristic differ from the ethics of persons having another characteristic. Self-selection bias occurs if people have control over that characteristic. When there is self-selection bias, we cannot be sure observed differences in ethics are correlated with the characteristic or are the result of individual self-selection. Self-selection bias is germane to many important business ethics questions. In this paper I explain what self- (...) class='Hi'>selection bias is, how it relates to business ethics research, and how to correct for it. I also illustrate the correction process in an empirical analysis of the effect of organizational rank on worker ethics. Using data from the European Values Survey, I find that being a supervisor is positively correlated with worker ethics. However, I also find a negative self-selection effect. Workers with relatively lower ethics are selected into supervisory roles. (shrink)

The prevalence, course and prognosis of diseases in patients referred to tertiary medical centers frequently differ from those treated in primary care settings. Extrapolation of findings from one population to another may therefore be unwarranted. Other factors that contribute to misinterpretation of medical literature include failure to distinguish statistical from clinical significance and advocacy of medical interventions prior to adequate clinical trials.

Observations with respect to the relationship between symptoms and diseases can seriously be biased by selection phenomena. This selection may occur from the general population, via consultation behavior, diagnostic and therapeutic activities of the general practitioner, and by referral.Relationships may be suggested and reproduced even if they do not exist in unselected populations, as a product of diagnostic routines. Correction for selection bias can only be achieved by choosing proper comparison groups. While this can be done (...) in a general practice setting, this is almost impossible after referral, as is demonstrated in this paper. Surprisingly, the most unbiased estimation of the relationship between symptoms and diseases after referral can be made from patient groups that are referred for reason unrelated to the disease under study. (shrink)

Philosophers of science have argued that epistemic diversity is an asset for the production of scientific knowledge, guarding against the effects of biases, among other advantages. The growing privatization of scientific research, on the contrary, has raised important concerns for philosophers of science, especially with respect to the growing sources of biases in research that it seems to promote. Recently, Holman and Bruner ( 2017 ) have shown, using a modified version of Zollman ( 2010 ) social network model, that (...) an industrial selection bias can emerge in a scientific community, without corrupting any individual scientist, if the community is epistemically diverse. In this paper, we examine the strength of industrial selection using a reinforcement learning model, which simulates the process of industrial decision-making when allocating funding to scientific projects. Contrary to Holman and Bruner’s model, in which the probability of success of the agents when performing an action is given a priori, in our model the industry learns about the success rate of individual scientists and updates the probability of success on each round. The results of our simulations show that even without previous knowledge of the probability of success of an individual scientist, the industry is still able to disrupt scientific consensus. In fact, the more epistemically diverse the scientific community, the easier it is for the industry to move scientific consensus to the opposite conclusion. Interestingly, our model also shows that having a random funding agent seems to effectively counteract industrial selection bias. Accordingly, we consider the random allocation of funding for research projects as a strategy to counteract industrial selection bias, avoiding commercial exploitation of epistemically diverse communities. (shrink)

Context: Political psychology rests on the assumption of the existence of a world outside and independent of consciousness. This ontological materialism is hardly spoken of within the field, as it ….

This work examines the influence of reference dependence, including value selection bias and congruence effects, on diagnostic reasoning. Across two studies, we explored how dependence on the initial problem structure influences the ability to solve simplified precursors to the more traditional Bayesian reasoning problems. Analyses evaluated accuracy and types of response errors as a function of congruence between the problem presentation and question of interest, amount of information, need for computation, and individual differences in numerical abilities. Across all (...) problem variations, there was consistent and strong evidence of a value selection bias in that incorrect responses almost always conformed to values that were provided in the problem rather than other errors including those related to computation. The most consistent and unexpected error across all conditions in the first experiment was that people were often more likely to utilize the superordinate value as part of their solution rather than the anticipated reference class values. This resulted in a weakened effect of congruence, with relatively low accuracy even in congruent conditions, and a dominant response error of the superordinate value. Experiment 2 confirmed that the introduction of a new sample drew attention away from the provided reference class, increasing reliance on the overall sample size. This superordinate preference error, along with the benefit of repeating the PPV reference class within the question, demonstrated the importance of reference dependence based on the salience of information within the response prompt. Throughout, higher numerical skills were generally associated with higher accuracy, whether calculations were required or not. (shrink)

Whenever the use of non-experimental data for discovering causal relations or predicting the outcomes of experiments or interventions is contemplated, two difficulties are routinely faced. One is the problem of latent variables, or confounders: factors influencing two or more measured variables may not themselves have been measured or recorded. The other is the problem of sample selection bias: values of the variables or features under study may themselves influence whether a unit is included in the data sample.

Objectives: To investigate whether different methods of obtaining informed consent affected recruitment to a study of delirium in older, medically ill hospital inpatients.Design: Open randomised study.Setting: Acute medical service for older people in an inner city teaching hospital.Participants: Patients 70 years or older admitted to the unit within three days of hospital admission randomised into two groups.Intervention: Attempted recruitment of subjects to a study of the natural history of delirium. This was done by either a formal test of capacity, followed (...) by either a request for consent or an attempt at obtaining assent from a proxy, or a combined informal capacity/consent process.Main outcome measures: Prevalence and severity of delirium, and, as case mix measures, length of hospital stay and destination on discharge.Results: Recruitment of subjects through establishing formal capacity and then informed consent was less successful and, compared with those recruited through the usual combined capacity/consent approach, yielded a sample with less cognitive impairment, lower severity of delirium, lower probability of case note diagnosis of delirium and lower rate of entering a care home.Conclusions: Methods of obtaining informed consent may significantly influence the case mix of subjects recruited to a study of delirium. Stringent testing of capacity may exclude patients with delirium from studies, thus rendering findings less generalisable. A different method is necessary to achieve an ethical balance between respecting autonomy through obtaining adequate informed consent and avoiding sample bias. (shrink)

Causal discovery becomes especially challenging when the possibility of latent confounding and/or selection bias is not assumed away. For this task, ancestral graph models are particularly useful in that they can represent the presence of latent confounding and selection effect, without explicitly invoking unobserved variables. Based on the machinery of ancestral graphs, there is a provably sound causal discovery algorithm, known as the FCI algorithm, that allows the possibility of latent confounders and selection bias. However, (...) the orientation rules used in the algorithm are not complete. In this paper, we provide additional orientation rules, augmented by which the FCI algorithm is shown to be complete, in the sense that it can, under standard assumptions, discover all aspects of the causal structure that are uniquely determined by facts of probabilistic dependence and independence. The result is useful for developing any causal discovery and reasoning system based on ancestral graph models. (shrink)

_Anthropic Bias_ explores how to reason when you suspect that your evidence is biased by "observation selection effects"--that is, evidence that has been filtered by the precondition that there be some suitably positioned observer to "have" the evidence. This conundrum--sometimes alluded to as "the anthropic principle," "self-locating belief," or "indexical information"--turns out to be a surprisingly perplexing and intellectually stimulating challenge, one abounding with important implications for many areas in science and philosophy. There are the philosophical thought experiments and (...) paradoxes: the Doomsday Argument; Sleeping Beauty; the Presumptuous Philosopher; Adam & Eve; the Absent-Minded Driver; the Shooting Room. And there are the applications in contemporary science: cosmology ; evolutionary theory ; the problem of time's arrow ; quantum physics ; game-theory problems with imperfect recall ; even traffic analysis. _Anthropic Bias_ argues that the same principles are at work across all these domains. And it offers a synthesis: a mathematically explicit theory of observation selection effects that attempts to meet scientific needs while steering clear of philosophical paradox. (shrink)

if and only if for every W in V, W is independent of the set of all its non-descendants conditional on the set of its parents. One natural question that arises with respect to DAGs is when two DAGs are “statistically equivalent”. One interesting sense of “statistical equivalence” is “d-separation equivalence” (explained in more detail below.) In the case of DAGs, d-separation equivalence is also corresponds to a variety of other natural senses of statistical equivalence (such as representing the same (...) set of distributions). Theorems characterizing d-separation equivalence for directed acyclic graphs and that can be used as the basis for polynomial time algorithms for checking d-separation equivalence were provided by Verma and Pearl (1990), and Frydenberg (1990). The question we will examine is how to extend these results to cases where a DAG may have latent (unmeasured) variables or selection bias (i.e. some of the variables in the DAG have been conditioned on.) D-separation equivalence is of interest in part because there are algorithms for constructing DAGs with latent variables and selection bias that are based on observed conditional independence relations. For this class of algorithms, it is impossible to determine which of two d-separation equivalent causal structures generated a given probability distribution, given only the set of conditional independence and dependence relations true of the observed distribution. We will describe a polynomial (in the number of vertices) time algorithm for determining when two DAGs which may have latent variables or selection bias are d-separation equivalent. (shrink)

The Berkson effect shows that two independent diseases, A and B, become negatively correlated if they are confined within the walls of a hospital. We explain that, simply by adding a third disease, C, the negative correlation may flip into a positive one, and we identify the point where this happens. That leads to a necessary and sufficient condition for a positive as well as a negative correlation between A and B. We further explain that a flip from negative to (...) positive is impossible if C is independent of A, of B, and of the disjunction of A and B: with these three independences in place, the Berkson effect remains in force. However, if only two of the three independences hold, the effect is not guaranteed. (shrink)

Today the modern Darwinian theory of evolution is the unifying theory within the biological sciences. A consideration of its implications for feminism is, however, impossible without a critical evaluation of its history of male bias. The aim of this article is therefore threefold. First, to explain what sexual selection entails. Second, to discuss male bias in and feminist reactions to Darwinian theory in general and sexual selection theory in particular. Third, to demonstrate that it would be (...) a loss for feminism to keep rejecting an evolutionary framework if we want to understand the roots of gender difference. This article is informed by a Darwinian feminist perspective. (shrink)

In a recent issue of this journal, Vandermassen suggested that feminists should include sexual selection theory and evolutionary psychology in a unifying theory of human nature. In response, this article aims to offer some insight into the development of sexual selection theory, to caution against Vandermassen's unreserved assimilation and to promote the opposite ongoing integration — an inclusion of gender perspectives into evolutionary biology. In society today, opinions about maintaining traditional sex roles are often put forward on the (...) basis of what is natural and how animals behave. However, the natural sciences have proved to be pervaded by gendered values and interests; Darwin's theory of sexual selection has been criticized for being male biased, and partly due to the unwillingness of Darwin's scientific contemporaries to accept female choice, research has been overwhelmingly focused on males. More recently, theory has become less gender biased and research has come to include a large variety of issues not present in the first version of the theory. However, there is a need to increase the awareness of gender bias in order to develop a gender-neutral evolutionary biology. (shrink)

Davey reviews evidence purporting to distinguish between two accounts of selective associations – expectancy bias and evolved predispositions, although these hypotheses largely apply to different levels of causal analysis. Criticisms of primate studies in which subjects lack prior exposure to stimuli seem uncompelling. Expectancies may sometimes serve as proximal mediators in selective associations, but other factors, both proximate and ultimate, are clearly also involved.

We tested the hypothesis that choices determined by Type 1 processes are compelling because they are fluent, and for this reason they are less subject to analytic thinking than other answers. A total of 104 participants completed a modified version of Wason's selection task wherein they made decisions about one card at a time using a two-response paradigm. In this paradigm participants gave a fast, intuitive response, rated their feeling of rightness for that response, and were then allowed free (...) time to reconsider their answers. As we predicted, answers consistent with a matching heuristic were made more quickly than other answers, were given higher FOR ratings, and received less subsequent analysis as measured by rethinking time and the probability of changing answers. These data suggest that reasoning biases may be compelling because they are fluently generated; this is turn creates a strong FOR, which acts as a signal that further analysis is not necessary. (shrink)

The processes that guide performance in Wason's selection task (WST) are still under debate. The matching bias effect in the negations paradigm and its elimination by explicit negations are central arguments against a substantial role for inferential processes. Two WST experiments were conducted in the negations paradigm to replicate the basic finding and to compare effects of implicit and explicit negations. Results revealed robust matching bias in implicit negations. In contrast to previous findings, matching bias was (...) reduced but not eliminated in conditions using explicit negations. Model-based analyses suggest that matching bias is due to a switch towards a negative test strategy caused by negations. (shrink)

The vast majority of extant empirical research examining the relationship between corporate social performance and financial performance selects samples of only those firms which are observed engaging in CSP. In this study, the authors assert that firms’ efforts to pursue CSP and subsequently their appearance in social-choice investment advisory firms’ ranking databases are non-random. Studying the CSP–FP link using selected samples of only those firms whose social performance is ranked by SIA firms introduces a sample-selection bias which limits (...) generalization of results to a population of all firms, and at worst provides alternate explanations for observed relationships. The authors test these assertions on a large sample of public corporations in the United States over 6 years and find a sample-selection bias. Upon correction of this bias, this study confirms the positive impact of CSP on FP. (shrink)

Animal welfare scientific literature has accumulated rapidly in recent years, but bias may exist which influences understanding of progress in the field. We conducted a survey of articles related to animal welfare or well being from an electronic database. From 8,541 articles on this topic, we randomly selected 115 articles for detailed review in four funding categories: government; charity and/or scientific association; industry; and educational organization. Ninety articles were evaluated after unsuitable articles were rejected. The welfare states of animals (...) in new treatments, conventional treatments or control groups with no treatment were classified as high, medium or low according to one or more. More articles were published in which the welfare of animals in new treatments was better than that of animals in the conventional or no treatment groups, demonstrating a positive result bias. Failure to publish studies with negative or inconclusive results may lead to other scientists unnecessarily repeating the research. The authors’ assessments of the welfare state of the groups were similarly rated high, medium or low, and it was found that new treatments were rated lower if the research was funded by industry, and higher when funded by charities, with government funding agencies intermediate. These differences were not evident in the Five Freedoms assessment, demonstrating an authors’ assessment bias that appeared to support the funding agencies’ interests. North American funded publications rated the welfare of animals in New treatments higher and those in a Conventional or No Treatment lower, compared with European-funded publications. It is concluded that preliminary evidence was provided of several forms of publication bias in animal welfare science. (shrink)

A cognitive bias is a pattern of deviation in our judgment or our processing of what we perceive. Its raison d'être is the evolutionary need to produce immediate judgments in order to adopt a position quickly in response to stimuli, problems, or situations that catch our attention for some reason. They have a social dimension because they are present in the interactions and decision-making processes of ordinary life. They can be understood to be an adaptive response to human inability (...) to process all of the available information that we receive and selectively filter through biased reasoning or through procedures and heuristic rules. (shrink)

The phenomenon known as matching bias consists of a tendency to see cases as relevant in logical reasoning tasks when the lexical content of a case matches that of a propositional rule, normally a conditional, which applies to that case. Matching is demonstrated by use of the negations paradigm that is by using conditionals in which the presence and absence of negative components is systematically varied. The phenomenon was first published in 1972 and the present paper reviews the history (...) of research and theorising on the problem in the subsequent 25 years. Theories of matching bias considered include those based on several broad frameworks including the heuristic-analytic theory, the mental models theory, the theory of optimal data selection, and relevance theory as well as the specific processing-negations account. The ability of these theories to account for a range of phenomena is considered, including the effects of linguistic form, realistic content, and explicit negation on the matching bias effect. Of particular importance are recent findings showing that the bias is observable on a wider range of linguistic forms than has generally been thought, and that it is almost entirely dependent on the use of implicit negation in the logical cases to which rules are applied. The reasons for the general suppression of matching when realistic content is used are, however, unclear and a need for further research is identified here. It is concluded that matching bias is a highly robust effect which is closely connected with the problem of understanding implicit negation. Most of the theories in the literature are unable to account for at least some of the major phenomena discovered in research on the bias. The accounts that fare best are those that posit local effects of negation, including the heuristic-analytic and processing negations theories. (shrink)

In this introduction, we acquaint readers with a selection of work coming out of the "Bias in Context" conference series, which ran from 2016 to 2017. Featured authors in the symposium include Gabriella Beckles-Raymond (writing about bad faith and implicit bias explanations), Daniel Kelly and Lacey Davidson (writing about gender norms and the internalization of social structures), and Alex Madva (writing about solutions to racial integration and an empirical mindset). We sketch the larger themes of the conference, (...) as well as how these three pieces of work fit into them. (shrink)

Women’s underrepresentation in science, technology, engineering, and mathematics (STEM) is related to the hierarchical social structure of gender relations in these fields. However, interventions to increase women’s participation have focused primarily on women’s interests rather than on STEM managers’ hiring practices. In this research, we examine STEM hiring practices, explore the implicit bias in criteria used by STEM managers, and suggest possible corrective solutions. Using an experimental design with 213 men and women STEM managers, we show that when evaluating (...) a female candidate, women and men STEM managers apply differential selection criteria, with men demonstrating implicit in-group gender favoritism in their hiring decisions. Specifically, the ability to work long hours was a more important criterion for male managers when evaluating a female candidate, forming an implicit gender bias, whereas female managers gave greater importance to problem-solving ability, a more gender-equal criterion. Adding a personal note to the curriculum vitae stating that the candidate had hired a full-time nanny was useful in decreasing the importance of the ability to work long hours criterion for men managers. We suggest individual and institutional interventions to reduce this bias, as a path to increasing women’s participation in STEM. (shrink)

Animal welfare scientific literature has accumulated rapidly in recent years, but bias may exist which influences understanding of progress in the field. We conducted a survey of articles related to animal welfare or well being from an electronic database. From 8,541 articles on this topic, we randomly selected 115 articles for detailed review in four funding categories: government; charity and/or scientific association; industry; and educational organization. Ninety articles were evaluated after unsuitable articles were rejected. The welfare states of animals (...) in new treatments, conventional treatments or control groups with no treatment were classified as high, medium or low according to one or more. More articles were published in which the welfare of animals in new treatments was better than that of animals in the conventional or no treatment groups, demonstrating a positive result bias. Failure to publish studies with negative or inconclusive results may lead to other scientists unnecessarily repeating the research. The authors’ assessments of the welfare state of the groups were similarly rated high, medium or low, and it was found that new treatments were rated lower if the research was funded by industry, and higher when funded by charities, with government funding agencies intermediate. These differences were not evident in the Five Freedoms assessment, demonstrating an authors’ assessment bias that appeared to support the funding agencies’ interests. North American funded publications rated the welfare of animals in New treatments higher and those in a Conventional or No Treatment lower, compared with European-funded publications. It is concluded that preliminary evidence was provided of several forms of publication bias in animal welfare science. (shrink)

Current research in artificial intelligence (AI) sheds light on algorithmic bias embedded in AI systems. The underrepresentation of women in the AI design sector of the tech industry, as well as in training datasets, results in technological products that encode gender bias, reinforce stereotypes and reproduce normative notions of gender and femininity. Biased behaviour is notably reflected in anthropomorphic AI systems, such as personal intelligent assistants (PIAs) and chatbots, that are usually feminized through various design parameters, such as (...) names, voices and traits. Gendering of AI entities, however, is often reduced to the encoding of stereotypical behavioural patterns that perpetuate normative assumptions about the role of women in society. The impact of this behaviour on social life increases, as human-to-(anthropomorphic)machine interactions are mirrored in human-to-human social interactions. This article presents current critical research on AI bias, focusing on anthropomorphic systems. Moreover, it discusses the significance of women’s engagement in AI design and programming, by presenting selected case studies of contemporary female media artists and designers. Finally, it suggests that women, through their creative practice, provide feminist and critical approaches to AI design which are essential for imagining alternative, inclusive, ethic and de-biased futures for anthropomorphic AIs. (shrink)

Phillips & Silverstein emphasize the gain-control properties of NMDA synapses in cognitive coordination. We endorse their view and suggest that NMDA synapses play a crucial role in biased attentional competition and (visual) working memory. Our simulations show that NMDA synapses can control the storage rate of visual objects. We discuss specific predictions of our model about cognitive effects of NMDA-antagonists and schizophrenia.

This book presents and deconstructs the existing explanations for the differential career development of qualified men and women. It reframes the problem of discrimination in the workplace as a matter of organizational ethics, social responsibility and compliance with existing equal opportunity laws. Sensitive points are identified where social biases, decision-makers' individual economic interests and shortcomings of organizational incentive policies may lead to discrimination against qualified women. The ideas put forward are empirically tested in an original laboratory experiment that examines personnel (...) selection in the male-dominated field of science and technology. It contrasts the selection of applicants with gendered and gender-blind applications available to subjects under controlled conditions. 30% of participants were high-level decision-makers, which is unprecedented in this field of research. The results, highly relevant for organizational practice, are explained and discussed in detail. (shrink)

The purpose of this research is to empirically test the salmon bias hypothesis, which states that the “healthy migrant” effect—referring to a situation in which migrants enjoy lower mortality risks than natives—is caused by selective return-migration of the weak, sick, and elderly. Using a unique longitudinal micro-level database—the Historical Sample of the Netherlands—we tracked the life courses of internal migrants after they had left the city of Rotterdam, which allowed us to compare mortality risks of stayers, returnees, and movers (...) using survival analysis for the study group as a whole, and also for men and women separately. Although migrants who stayed in the receiving society had significantly higher mortality risks than natives, no significant difference was found for migrants who returned to their municipality of birth. By contrast, migrants who left for another destination had much lower mortality risks than natives. Natives who left Rotterdam also had significantly lower mortality risks than natives who stayed in Rotterdam. Female migrants, in particular, who stayed in the receiving urban society paid a long-term health price. In the case of Rotterdam, the salmon bias hypothesis can be rejected because the lower mortality effect among migrants was not caused by selective return-migration. The healthy migrant effect is real and due to a positive selection effect: Healthier people are more likely to migrate. (shrink)

Researchers increasingly rely on large data sets of health information, often linked with biological specimens. In recent years, the argument has been made that obtaining informed consent for conducting records-based research is unduly burdensome and results in ?consent bias.? As a type of selection bias, consent bias is said to exist when the group giving researchers access to their data differs from the group denying access. Therefore, to promote socially beneficial research, it is argued that consent (...) should be unnecessary. After analyzing the biostatistics evidence and bioethics arguments, the article concludes that (1) claims about the amount of consent bias are overstated; (2) commonly used statistical methods usually can account for consent bias; and (3) any residual effects of consent bias are below an acceptable level of imprecision and constitute a reasonable social cost for conducting ethically responsible research. (shrink)

Two experiments were conducted to show that the IF … THEN … rules used in the different versions of Wason 's selection task are not psychologically—though they are logically—equivalent. Some of these rules are considered by the participants as strict logical conditionals, whereas others are interpreted as expressing a biconditional relationship. A deductive task was used jointly with the selection task to show that the original abstract rule is quite ambiguous in this respect, contrary to deontic rules: the (...) typical “error” made by most people may indeed be explained by the fact that they consider the abstract rule as a biconditional. Thus, there is no proper error or bias in the selection task as it is still argued, but a differential interpretation of the rule. The need for taking into account a pragmatic component in the process of reasoning is illustrated by the experiments. (shrink)

Neuroscience research, like all science, is vulnerable to the influence of extraneous values in the practice of research, whether in research design or the selection, analysis and interpretation of data. This is particularly problematic for research into the biological mechanisms that underlie behavior, and especially the neurobiological underpinnings of moral development and ethical reasoning, decision-making and behavior, and the other elements of what is often called the neuroscience of ethics. The problem arises because neuroscientists, like most everyone, bring to (...) their work assumptions, preconceptions and values and other sources of potentially inappropriate bias of which they may be unaware. It is important that the training of neuroscientists, and research practice itself, include open and in-depth discussion and examination of the assumptions that underlie research. Further, policy makers, journalists, and the general public, that is, the consumers of neuroscience research findings (and by extension, neurotechnologies) should be made aware of the limitations as well as the strengths of the science, the evolving nature of scientific understanding, and the often invisible values inherent in science. (shrink)

An experiment is reported which tests for positive confirmation bias in a setting in which individuals choose what information to buy, prior to making a decision. The design – an adaptation of Wason's selection task – reveals the use that subjects make of information after buying it. Strong evidence of positive confirmation bias, in both information acquisition and information use, is found; and this bias is found to be robust to experience. It is suggested that the (...) bias results from a pattern of reasoning which, although producing sub-optimal decisions, is internally coherent and which is self-reinforcing. (shrink)

Some cultural evolution researchers have argued for the importance of prestige bias as a systematic and widespread social learning bias, that structures human social learning and cultural transmission patterns. Broadly speaking, prestige bias accounts understand it as a bias towards copying ‘prestigious’ individuals. Prestige bias, along with other social learning biases, has been argued to pay a crucial role in allowing cumulative cultural selection to take place, thereby generating adaptations that are key to our (...) success as a species. However, I argue for skepticism about the plausibility and scope of a prestige bias account. I argue that although an account of prestige bias seems plausible or compelling on their face, it is committed to a particular view of the cognition underpinning the bias, and therefore to predictions regarding its flexibility and context-sensitivity. Given this, current empirical evidence gives us reason to doubt the explanatory value of a prestige bias account over a naive, goal-directed agent account. Additionally, the way that prestige is defined in empirical work is in tension with a general understanding of prestige, casting doubt upon its status as evidence of prestige bias. I examine two studies cited as evidence of prestige bias, arguing that in these cases we cannot clearly favour a prestige bias explanation over a goal-directed agent explanation. (shrink)

In this paper the arguments for optimal data selection and the contrast class account of negations in the selection task and the conditional inference task are summarised, and contrasted with the matching bias approach. It is argued that the probabilistic contrast class account provides a unified, rational explanation for effects across these tasks. Moreover, there are results that are only explained by the contrast class account that are also discussed. The only major anomaly is the explicit negations (...) effect in the selection task (Evans, Clibbens, & Rood, 1996), which it is argued may not be the result of normal interpretative processes. It is concluded that the effects of negation on human reasoning provide good evidence for the view that human reasoning processes may be rational according to a probabilistic standard. (shrink)

BackgroundPublication and outcome reporting bias is often caused by researchers selectively choosing which scientific results and outcomes to publish. This behaviour is ethically significant as it distorts the literature used for future scientific or clinical decision-making. This study investigates the practicalities of using ethics applications submitted to a UK National Health Service research ethics committee to monitor both types of reporting bias.MethodsAs part of an internal audit we accessed research ethics database records for studies submitting an end of (...) study declaration to the Hampshire A research ethics committee between 1st January 2010 and 31st December 2011. A literature search was used to establish the publication status of studies. Primary and secondary outcomes stated in application forms were compared with outcomes reported in publications.ResultsOut of 116 studies the literature search identified 57 publications for 37 studies giving a publication rate of 32 %. Original Research Ethics Committee applications could be obtained for 28 of the published studies. Outcome inconsistencies were found in 16 of the published studies.ConclusionsThis study showed that the problem of publication and outcome reporting bias is still significant in the UK. The method described here demonstrates that UK NHS research ethics committees are in a good position to detect such bias due to their unique access to original research protocols. Data gathered in this way could be used by the Health Research Authority to encourage higher levels of transparency in UK research. (shrink)