CiteULike: Tag bias

A novel, fast entropy-minimization algorithm for bias field correction in MR images

Qing Ji — 2007-07-26T12:37:23-00:00

Magnetic Resonance Imaging, Vol. 25, No. 2. (February 2007), pp. 259-264.

A novel, fast entropy-minimization algorithm for bias field correction in magnetic resonance (MR) images is suggested to correct the intensity inhomogeneity degradation of MR images that has become an increasing problem with the use of phased-array coils. Four important modifications were made to the conventional algorithm: (a) implementation of a modified two-step sampling strategy for stacked 2D image data sets, which included reducing the size of the measured image on each slice with a simple averaging method without changing the number of slices and then using a binary mask generated by a histogram threshold method to define the sampled voxels in the reduced image; (b) improvement of the efficiency of the correction function by using a Legendre polynomial as an orthogonal base function polynomial; (c) use of a nonparametric Parzen window estimator with a Gaussian kernel to calculate the probability density function and Shannon entropy directly from the image data; and (d) performing entropy minimization with a conjugate gradient method. Results showed that this algorithm could correct different types of MR images from different types of coils acquired at different field strengths very efficiently and with decreased computational load.

Point and Interval Estimation of Primary and Secondary Parameters in a Two-Stage Adaptive Clinical Trial

Aiyi Liu — 2008-03-14T21:45:59-00:00

Journal of Biopharmaceutical Statistics, Vol. 18, No. 2. (2008), pp. 211-226.

Investigated in this paper is the point estimation and confidence intervals of the treatment efficacy parameter and related secondary parameters in a two-stage adaptive trial. Based on the minimal sufficient statistics, several alternative estimators to the sample averages are proposed to reduce the bias and to improve the precision of estimation. Confidence intervals are constructed using Woodroofe's pivot method. Numerical studies are conducted to evaluate the bias and mean squared error of the estimators and the coverage probability of the confidence intervals.

Effect of interpretive bias on research evidence.

TJ Kaptchuk — 2006-11-09T20:13:07-00:00

BMJ, Vol. 326, No. 7404. (28 June 2003), pp. 1453-1455.

Response to the Letter from

Iain Purchase — 2007-06-27T21:06:38-00:00

Toxicology, Vol. 212, No. 2-3. (1 September 2005), pp. 253-254.

The Interplay of Science and Values in Assessing and Regulating Environmental Risks

Frances Lynn — 2006-11-09T20:03:48-00:00

Science, Technology, & Human Values, Vol. 11, No. 2. (1986), pp. 40-50.

Publication prejudices: An experimental study of confirmatory bias in the peer review system

Michael Mahoney — 2006-11-09T20:02:20-00:00

Cognitive Therapy and Research, Vol. V1, No. 2. (1 June 1977), pp. 161-175.

Biases about Man-made Cancer among Researchers

Allan Mazur — 2006-11-09T19:51:51-00:00

Social Studies of Science, Vol. 31, No. 5. (1 October 2001), pp. 771-778.

Can sociological variables significantly predict researchers' positions on a scientific issue? Carcinogenesis offers a good test case because it is not definitively established how important man-made substances are as contributors to human cancer. American cancer researchers divide on this issue, some broadly emphasizing the importance of man-made carcinogens, others consistently minimizing their importance relative to natural carcinogens. We find that cancer researchers who are male, are affiliated with a medical school, and have more publications, are more likely than their colleagues to minimize the importance of man-made carcinogens. 10.1177/030631201031005004

175th anniversary lecture. Medical journals and the shaping of medical knowledge.

JP Vandenbroucke — 2006-11-09T19:42:57-00:00

Lancet, Vol. 352, No. 9145. (6 1998), pp. 2001-2006.

Science, Biases, and the Threat of Global Pessimism

Brad Wray — 2006-11-09T19:37:21-00:00

Philip Kitcher rejects the global pessimists' view that the conclusions reached in inquiry are determined by the interests of some segment of the population, arguing that only some inquiries, for example, inquiries into race and gender, are adversely affected by interests. I argue that the biases Kitcher believes affect such inquiries are operative in all domains, but the prevalence of such biases does not support global pessimism. I argue further that in order to address the global pessimists' concerns, the scientific community needs criticism from people with diverse interests and background assumptions.

Scientific Judgment and the Limits of Conflict-of-Interest Policies

Kevin Elliott — 2008-01-07T14:28:15-00:00

Accountability in Research, Vol. 15, No. 1. (2008), pp. 1-29.

This article argues that the three major elements of typical university conflict-of-interest (COI) policies (i.e., disclosure, management, and elimination of conflicts via divestiture or recusal) are likely to be insufficient for screening out many worrisome influences of financial COIs. Current psychological research challenges the effectiveness of disclosure, management plans are unlikely to address the wide range of ways that financial COIs can influence scientific judgment, and it is often impractical to eliminate conflicts. Identifying the limits of these policies highlights the importance of considering alternative strategies, such as encouraging more independently funded research, in order to maintain the integrity of science.

Science, Values and Objectivity

Peter Davson-Galle — 2006-11-09T18:49:05-00:00

Science & Education, Vol. V11, No. 2. (1 March 2002), pp. 191-202.

Whose evidence? Lessons from the philosophy of science and the epistemology of medicine.

E Harari — 2007-03-17T13:52:38-00:00

Aust N Z J Psychiatry, Vol. 35, No. 6. (December 2001), pp. 724-730.

OBJECTIVE: To demonstrate the inadequacies of empiricism as a scientific foundation for evidence-based approaches to psychiatry. METHOD: The principles of empiricism are reviewed in the light of developments in the philosophy of science and phenomenology. Case studies are selected from the history of physical sciences, biological science and clinical sciences (pathology, neuroscience, psychosocial science and psychopathology), paying particular attention to the role of observation in theory construction. RESULTS: The principles of empiricism, particularly its view of the nature of observation as the basis of evidence do not reflect the historical reality of scientific theorizing and practice. Science has constructed alternative models of its own activity that do justice to the complexities of its subject matter, including the world of human experience and mental illness. CONCLUSIONS: A failure to recognize both the limitations of empiricism in science and the conceptual richness of alternative formulations that accord more closely with the complexity of psychiatry's domain will result in a naïve model of science and inadequate understanding of the limitations of 'evidence' that guide the training, clinical practice and research in our profession. The consequences will be the intellectual, clinical and ethical impoverishment of psychiatry.

The past suppression of industry knowledge of the toxicity of benzene to humans and potential bias in future benzene research.

PF Infante — 2007-06-19T13:56:18-00:00

Int J Occup Environ Health, Vol. 12, No. 3. (p 2006), pp. 268-272.

Petrochemical industry representatives often withhold information and misinterpret positive evidence of toxicity of benzene, even from their own research, also discouraging or delaying disclosure of findings of adverse effects to the public. They now appear to be attempting to influence study results in industry's favor by offering predetermined conclusions about study results as part of an effort to draw financial support for the studies. The American Petroleum Institute is currently raising funds for benzene research being conducted in China for which it has already announced the intended conclusions.

How the choice of experimental organism matters: epistemological reflections on an aspect of biological practice.

RM Burian — 2007-03-15T21:48:10-00:00

J Hist Biol, Vol. 26, No. 2. (1993), pp. 351-367.

Ten categories of statistical errors: a guide for research in endocrinology and metabolism.

TH Holmes — 2007-12-21T15:54:31-00:00

Am J Physiol Endocrinol Metab, Vol. 286, No. 4. (April 2004)

A simple framework is introduced that defines ten categories of statistical errors on the basis of type of error, bias or imprecision, and source: sampling, measurement, estimation, hypothesis testing, and reporting. Each of these ten categories is illustrated with examples pertinent to research and publication in the disciplines of endocrinology and metabolism. Some suggested remedies are discussed, where appropriate. A review of recent issues of American Journal of Physiology: Endocrinology and Metabolism and of Endocrinology finds that very small sample sizes may be the most prevalent cause of statistical error in this literature.

Publication Decisions and Their Possible Effects on Inferences Drawn from Tests of Significance--Or Vice Versa

Theodore Sterling — 2007-02-07T11:56:06-00:00

Journal of the American Statistical Association, Vol. 54, No. 285. (1959), pp. 30-34.

There is some evidence that in fields where statistical tests of significance are commonly used, research which yields nonsignificant results is not published. Such research being unknown to other investigators may be repeated independently until eventually by chance a significant result occurs-an "error of the first kind"-and is published. Significant results published in these fields are seldom verified by independent replication. The possibility thus arises that the literature of such a field consists in substantial part of false conclusions resulting from errors of the first kind in statistical tests of significance.

Missing data: prevalence and reporting practices.

TE Bodner — 2008-05-29T14:02:46-00:00

Psychological reports, Vol. 99, No. 3. (December 2006), pp. 675-680.

Results are described for a survey assessing prevalence of missing data and reporting practices in studies with missing data in a random sample of empirical research journal articles from the PsychINFO database for the year 1999, two years prior to the publication of a special section on missing data in Psychological Methods. Analysis indicates missing data problems were found in about one-third of the studies. Further, analytical methods and reporting practices varied widely for studies with missing data. One may consider these results as baseline data to assess progress as reporting standards evolve for studies with missing data. Some potential reporting standards are discussed.

Bias in toxicology.

B Wandall — 2007-08-08T12:49:21-00:00

Arch Toxicol, Vol. 81, No. 9. (September 2007), pp. 605-617.

The potential for bias, i.e., influences that cause results to deviate systematically from the truth is substantial both in toxicological research and in the performance of standardized toxicological testing. In this contribution, major potential sources of bias in toxicological research and testing are identified. Due to the lack of empirical studies of bias in toxicology, very little is known about its prevalence and impact. Areas to consider for such studies are pointed out, and it is suggested that such investigations should be given priority.

Large effects from small exposures. II. The importance of positive controls in low-dose research on bisphenol A

Vom — 2006-11-09T15:25:38-00:00

Environmental Research, Vol. 100, No. 1. (January 2006), pp. 50-76.

Over six-billion pounds per year of the monomer bisphenol A (BPA) are used to manufacture polycarbonate plastic products, resins lining cans, dental sealants, and polyvinyl chloride plastic products. There are 109 published studies as of July 2005 that report significant effects of low doses of BPA in experimental animals, with many adverse effects occurring at blood levels in animals within and below average blood levels in humans; 40 studies report effects below the current reference dose of 50 [mu]g/kg/day that is still assumed to be safe by the US-FDA and US-EPA in complete disregard of the published findings. The extensive list of significant findings from government-funded studies is compared to the 11 published studies that were funded by the chemical industry, 100% of which conclude that BPA causes no significant effects. We discuss the importance of appropriate controls in toxicological research and that positive controls are required to determine whether conclusions from experiments that report no significant effects are valid or false.

Understanding Bias in Scientific Practice

Nancy Shaffer — 2006-11-09T15:20:29-00:00

Methodological objectivism is a conception of bias which obscures the contingent and limited nature of methodological principles behind the guise of fixed a priori standards. I suggest as an alternative a more flexible view of the operation of bias which I call the attribution model. The attribution model makes explicit the working principles of both parties to an actual charge of bias. It enables those involved to identify the issues in dispute between them, and is the basis for an approach to handling charges of bias within the process of discussion and negotiation which characterizes normal scientific decision-making.

Fraud, errors and gamesmanship in experimental toxicology

Iain Purchase — 2006-11-09T15:14:06-00:00

Toxicology, Vol. 202, No. 1-2. (30 September 2004), pp. 1-20.

We expect moral behaviour from scientists. Morality implies being a good person and being good at one's profession. The general view appears to be that the vast majority of scientists aim to achieve these high standards.Science prides itself on the `self-correcting' mechanism in the scientific method, namely the requirement to reproduce findings before they are taken seriously. However, when findings are related to the adverse effects of chemicals there are several features that make this less effective than in some other fields of science. First, is the perception that everyone is exposed to chemicals and observations about chemical danger are immediately applicable to many people. Second, it is often easy to summarize adverse findings in attention-getting headlines seen by the lay public before the slow process of replication and interpretation has time to work. Third, most regulatory toxicology studies on a particular compound are only done once to minimise cost and the use of animals. Finally, the question posed about chemicals - are they safe? - is easy to ask but more difficult to test with appropriate studies.Fabrication of data in regulatory studies was found to occur in several contract laboratories in the 1960s and this lead directly to the introduction of Good Laboratory Practice regulations. Now studies submitted for regulatory purposes must comply with GLP regulations and this has virtually eliminated flawed studies due to fraudulent or careless behaviour.It is possible to discern different ways in which the expected standards have not been met. The first is in the intention of the work. Thus reports that the Roodeplaats Research Laboratory in South Africa was seeking to identify toxins that would kill without trace is an example where the intention is unacceptable. The second is in the conduct of the studies. Here the examples of William McBride and Michael Briggs who falsified data are pertinent. The example of the retraction of reports on the toxicity of ecstasy because the wrong compound had been administered indicates a degree of carelessness in the conduct of the study. The third is in the design and interpretation of studies. The report that genetic modification per se could render potatoes toxic has been criticised because of the inappropriate design and interpretation of the studies. Finally, that the reports of studies are biased because of conflicts of interest. Journals often require a declaration that the author has no financial conflict of interest. However, there are many other conflicts of interest with just as large an impact on the author's impartiality which are omitted from consideration. Gamesmanship has also entered the practice of toxicology, for example where strong assertions about conflict of interest are used to justify particular points of view.The main casualty from fraud, errors and gamesmanship is the perceived status of science itself. It is only gamesmanship that is on the increase. The remedies for these activities are explored.

Environmental risk assessment and the intrusion of bias

David Ball — 2006-11-09T15:01:56-00:00

Environment International, Vol. 28, No. 6. (December 2002), pp. 529-544.

The concept of managing the environment and any associated human health impacts by means of such science-based tools as toxicological evaluation, risk assessment, and economic appraisal has become widely accepted in professional circles. These increasingly complex methodologies have not won universal support, however, even among the technically minded, and the wider public has in many cases remained sceptical. The public's seeming lack of enthusiasm has frequently been assigned to ignorance of science, irrationality even, and some attempts have been made to ameliorate the situation by means of education, though with little evidence of success. However, this review advances an alternative explanation, namely, that the disenchantment has more to do with procedural than technical matters. Many issues, although treated with technical risk assessment, appear to have forced themselves onto the agenda because of factors only passingly connected with risk, and which are more related to the intrusion of disputed values and other sources of bias. This paper also notes that the intervention of bias in decision making is rife, and is found as much in professional as in public approaches. This need not itself be a problem, providing it is recognised and openly expressed. However, there remains a need for much greater circumspection and frankness by professions about the status of their art, and a determined effort if the full social benefits of environmental risk assessment (ERA) are to be realised.

Bias in analytic research.

DL Sackett — 2006-11-24T20:26:45-00:00

J Chronic Dis, Vol. 32, No. 1-2. (1979), pp. 51-63.

Publication bias: evidence of delayed publication in a cohort study of clinical research projects.

JM Stern — 2006-11-24T20:23:08-00:00

BMJ, Vol. 315, No. 7109. (13 September 1997), pp. 640-645.

OBJECTIVES: To determine the extent to which publication is influenced by study outcome. DESIGN: A cohort of studies submitted to a hospital ethics committee over 10 years were examined retrospectively by reviewing the protocols and by questionnaire. The primary method of analysis was Cox's proportional hazards model. SETTING: University hospital, Sydney, Australia. STUDIES: 748 eligible studies submitted to Royal Prince Alfred Hospital Ethics Committee between 1979 and 1988. MAIN OUTCOME MEASURES: Time to publication. RESULTS: Response to the questionnaire was received for 520 (70%) of the eligible studies. Of the 218 studies analysed with tests of significance, those with positive results (P < 0.05) were much more likely to be published than those with negative results (P > or = 0.10) (hazard ratio 2.32 (95% confidence interval 1.47 to 3.66), P = 0.0003), with a significantly shorter time to publication (median 4.8 v 8.0 years). This finding was even stronger for the group of 130 clinical trials (hazard ratio 3.13 (1.76 to 5.58). P = 0.0001), with median times to publication of 4.7 and 8.0 years respectively. These results were not materially changed after adjusting for other significant predictors of publication. Studies with indefinite conclusions (0.05 < or = P < 0.10) tended to have an even lower publication rate and longer time to publication than studies with negative results (hazard ratio 0.39 (0.13 to 1.12), P = 0.08). For the 103 studies in which outcome was rated qualitatively, there was no clear cut evidence of publication bias, although the number of studies in this group was not large. CONCLUSIONS: This study confirms the evidence of publication bias found in other studies and identifies delay in publication as an additional important factor. The study results support the need for prospective registration of trials to avoid publication bias and also support restricting the selection of trials to those started before a common date in undertaking systematic reviews.

Empirical evidence for selective reporting of outcomes in randomized trials: comparison of protocols to published articles.

AW Chan — 2006-05-26T12:00:18-00:00

JAMA, Vol. 291, No. 20. (26 May 2004), pp. 2457-2465.

CONTEXT: Selective reporting of outcomes within published studies based on the nature or direction of their results has been widely suspected, but direct evidence of such bias is currently limited to case reports. OBJECTIVE: To study empirically the extent and nature of outcome reporting bias in a cohort of randomized trials. DESIGN: Cohort study using protocols and published reports of randomized trials approved by the Scientific-Ethical Committees for Copenhagen and Frederiksberg, Denmark, in 1994-1995. The number and characteristics of reported and unreported trial outcomes were recorded from protocols, journal articles, and a survey of trialists. An outcome was considered incompletely reported if insufficient data were presented in the published articles for meta-analysis. Odds ratios relating the completeness of outcome reporting to statistical significance were calculated for each trial and then pooled to provide an overall estimate of bias. Protocols and published articles were also compared to identify discrepancies in primary outcomes. MAIN OUTCOME MEASURES: Completeness of reporting of efficacy and harm outcomes and of statistically significant vs nonsignificant outcomes; consistency between primary outcomes defined in the most recent protocols and those defined in published articles. RESULTS: One hundred two trials with 122 published journal articles and 3736 outcomes were identified. Overall, 50% of efficacy and 65% of harm outcomes per trial were incompletely reported. Statistically significant outcomes had a higher odds of being fully reported compared with nonsignificant outcomes for both efficacy (pooled odds ratio, 2.4; 95% confidence interval [CI], 1.4-4.0) and harm (pooled odds ratio, 4.7; 95% CI, 1.8-12.0) data. In comparing published articles with protocols, 62% of trials had at least 1 primary outcome that was changed, introduced, or omitted. Eighty-six percent of survey responders (42/49) denied the existence of unreported outcomes despite clear evidence to the contrary. CONCLUSIONS: The reporting of trial outcomes is not only frequently incomplete but also biased and inconsistent with protocols. Published articles, as well as reviews that incorporate them, may therefore be unreliable and overestimate the benefits of an intervention. To ensure transparency, planned trials should be registered and protocols should be made publicly available prior to trial completion.

Kinetic Bias in Estimates of Coastal Picoplankton Community Structure Obtained by Measurements of Small-Subunit rRNA Gene PCR Amplicon Length Heterogeneity

Marcelino Suzuki — 2008-05-11T01:05:46-00:00

Appl. Environ. Microbiol., Vol. 64, No. 11. (1 November 1998), pp. 4522-4529.

Marine bacterioplankton diversity was examined by quantifying natural length variation in the 5' domain of small-subunit (SSU) rRNA genes (rDNA) amplified by PCR from a DNA sample from the Oregon coast. This new technique, length heterogeneity analysis by PCR (LH-PCR), determines the relative proportions of amplicons originating from different organisms by measuring the fluorescence emission of a labeled primer used in the amplification reaction. Relationships between the sizes of amplicons and gene phylogeny were predicted by an analysis of 366 SSU rDNA sequences from cultivated marine bacteria and from bacterial genes cloned directly from environmental samples. LH-PCR was used to compare the distribution of bacterioplankton SSU rDNAs from a coastal water sample with that of an SSU rDNA clone library prepared from the same sample and also to examine the distribution of genes in the PCR products from which the clone library was prepared. The analysis revealed that the relative frequencies of genes amplified from natural communities are highly reproducible for replicate sets of PCRs but that a bias possibly caused by the reannealing kinetics of product molecules can skew gene frequencies when PCR product concentrations exceed threshold values.

Heteroduplexes in mixed-template amplifications: formation, consequence and elimination by 'reconditioning PCR'

Janelle Thompson — 2008-05-04T21:02:22-00:00

Nucl. Acids Res., Vol. 30, No. 9. (1 May 2002), pp. 2083-2088.

Although it has been recognized that PCR amplification of mixed templates may generate sequence artifacts, the mechanisms of their formation, frequency and potential elimination have not been fully elucidated. Here evidence is presented for heteroduplexes as a major source of artifacts in mixed-template PCR. Nearly equal proportions of homoduplexes and heteroduplexes were observed after co-amplifying 16S rDNA from three bacterial genomes and analyzing products by constant denaturing capillary electrophoresis (CDCE). Heteroduplexes became increasingly prevalent as primers became limiting and/or template diversity was increased. A model exploring the fate of cloned heteroduplexes during MutHLS-mediated mismatch repair in the Escherichia coli host demonstrates that the diversity of artifactual sequences increases exponentially with the number of both variable nucleotides and of original sequence variants. Our model illustrates how minimization of heteroduplex molecules before cloning may reduce artificial genetic diversity detected during sequence analysis by clone screening. Thus, we developed a method to eliminate heteroduplexes from mixed-template PCR products by subjecting them to reconditioning PCR', a low cycle number re-amplification of a 10-fold diluted mixed-template PCR product. This simple modification to the protocol may ensure that sequence richness encountered in clone libraries more closely reflects genetic diversity in the original sample. 10.1093/nar/30.9.2083

Biases in human estimation of interruptibility: effects and implications for practice

Daniel Avrahami — 2007-05-16T03:15:05-00:00

(2007), pp. 50-60.

Bihemispheric Leftward Bias in a Visuospatial Attention-Related Network

Tali Siman-Tov — 2007-12-19T03:47:04-00:00

J. Neurosci., Vol. 27, No. 42. (17 October 2007), pp. 11271-11278.

Asymmetry of spatial attention has long been described in both disease (hemispatial neglect) and healthy (pseudoneglect) states. Although right-hemisphere specialization for spatial attention has been suggested, the exact neural mechanisms of asymmetry have not been deciphered yet. A recent functional magnetic resonance imaging study from our laboratory serendipitously revealed bihemispheric left-hemifield superiority in activation of a visuospatial attention-related network. Nineteen right-handed healthy adult females participated in two experiments of visual half-field presentation. Either facial expressions (experiment 1) or house images (experiment 2) were presented unilaterally and parafoveally for 150 ms while subjects were engaging a central fixation task. Brain regions previously associated with a visuospatial attention network, in both hemispheres, were found to be more robustly activated by left visual field stimuli. The consistency of this finding with manifestations of attention lateralization is discussed, and a revised model based on neural connectivity asymmetry is proposed. Support for the revised model is given by a dynamic causal modeling analysis. Unraveling the basis for attention asymmetry may lead to better understanding of the pathogenesis of attention disorders, followed by improved diagnosis and treatment. Additionally, the proposed model for asymmetry of visuospatial attention might provide important insights into the mechanisms underlying functional brain lateralization in general. 10.1523/JNEUROSCI.0599-07.2007

A successful conversion or double refusal: A study of the process of refusal conversion in telephone survey research

Kana Fuse — 2007-10-26T07:25:49-00:00

The Social Science Journal, Vol. 44, No. 3. (2007), pp. 434-446.

This study investigated factors that affect the outcome of refusal conversion efforts in telephone surveys. We used data collected from three types of forms filled out by telephone interviewers: a Refusal Report Form was filled out when the respondent/household member refused to do the survey at the initial contact, and a Refusal Conversion Report Form or a Double Refusal Report Form was filled out during the refusal conversion phase. These forms included information such as the degree of interaction between interviewer and respondent/household member, reasons for refusal, information about the survey revealed during the introduction, and demographics. Results from logistic regressions suggested that the degree of interviewer-respondent/household member interaction during the second call attempt significantly influenced the outcome of refusal conversions. Results also suggested that revealing information about the legitimacy of the survey did not affect the outcome, and that there might even be a "boomerang effect" if interviewers rush to reveal such information. While respondent/household member's age did not affect the outcome, women appeared to be more difficult to be converted than men, especially when the degree of interviewer-respondent/household member interaction is low. Implication of these results and direction for future research directions are also discussed.

Measuring Media Bias: A Content Analysis of Time and Newsweek Coverage of Domestic Social Issues, 1975-2000*

Tawnya Covert — 2007-07-09T08:49:19-00:00

Social Science Quarterly, Vol. 88, No. 3. (2007), pp. 690-706.

Objective. This study is an effort to produce a more systematic, empirically-based, historical-comparative understanding of media bias than generally is found in previous works. Methods. The research employs a quantitative measure of ideological bias in a formal content analysis of the United States' two largest circulation news magazines, Time and Newsweek. Findings are compared with the results of an identical examination of two of the nation's leading partisan journals, the conservative National Review and the liberal Progressive. Results. Bias scores reveal stark differences between the mainstream and the partisan news magazines' coverage of four issue areas: crime, the environment, gender, and poverty. Conclusion. Data provide little support for those claiming significant media bias in either ideological direction.

Expert Political Judgment : How Good is It? How Can We Know?

Philip Tetlock — 2006-03-18T01:21:28-00:00

(05 July 2005)

The intelligence failures surrounding the invasion of Iraq dramatically illustrate the necessity of developing standards for evaluating expert opinion. This book fills that need. Here, Philip E. Tetlock explores what constitutes good judgment in predicting future events, and looks at why experts are often wrong in their forecasts. Tetlock first discusses arguments about whether the world is too complex for people to find the tools to understand political phenomena, let alone predict the future. He evaluates predictions from experts in different fields, comparing them to predictions by well-informed laity or those based on simple extrapolation from current trends. He goes on to analyze which styles of thinking are more successful in forecasting. Classifying thinking styles using Isaiah Berlin's prototypes of the fox and the hedgehog, Tetlock contends that the fox--the thinker who knows many little things, draws from an eclectic array of traditions, and is better able to improvise in response to changing events--is more successful in predicting the future than the hedgehog, who knows one big thing, toils devotedly within one tradition, and imposes formulaic solutions on ill-defined problems. He notes a perversely inverse relationship between the best scientific indicators of good judgement and the qualities that the media most prizes in pundits--the single-minded determination required to prevail in ideological combat. Clearly written and impeccably researched, the book fills a huge void in the literature on evaluating expert opinion. It will appeal across many academic disciplines as well as to corporations seeking to develop standards for judging expert decision-making.

Citation frequency: A biased measure of research impact significantly influenced by the geographical origin of research articles

Gerard Pasterkamp — 2007-11-22T07:02:10-00:00

Scientometrics, Vol. 70, No. 1. (15 January 2007), pp. 153-165.

Abstract Context. The use of citation frequency and impact factor as measures of research quality and journal prestige is being criticized. Citation frequency is augmented by self-citation and for most journals the majority of citations originate from a minority of papers. We hypothesized that citation frequency is also associated with the geographical origin of the research publication. Objective. We determined whether citations originate more frequently from institutes that are located in the same country as the authors of the cited publication than would be expected by chance. Design. We screened citations referring to 1200 cardiovascular publications in the 7 years following their publication. For the 1200 citation recipient publications we documented the country where the research originated (9 countries/regions) and the total number of received citations. For a selection of 8864 citation donor papers we registered the country/region where the citing paper originated. Results. Self-citation was common in cardiovascular journals (n = 1534, 17.8%). After exclusion of self-citation, however, the number of citations that originated from the same country as the author of the citation recipient was found to be on average 31.6% higher than would be expected by chance (p<0.01 for all countries/regions). In absolute numbers, nation oriented citation bias was most pronounced in the USA, the country with the largest research output (p<0.001). Conclusion. Citation frequency was significantly augmented by nation oriented citation bias. This nation oriented citation behaviour seems to mainly influence the cumulative citation number for papers originating from the countries with a larger research output.

Aid and bias

Keith Horton — 2004-12-28T17:23:58-00:00

Inquiry, Vol. 47, No. 6., 545.

Bias.

M Delgado-Rodríguez — 2007-08-26T04:37:37-00:00

J Epidemiol Community Health, Vol. 58, No. 8. (August 2004), pp. 635-641.

The concept of bias is the lack of internal validity or incorrect assessment of the association between an exposure and an effect in the target population in which the statistic estimated has an expectation that does not equal the true value. Biases can be classified by the research stage in which they occur or by the direction of change in a estimate. The most important biases are those produced in the definition and selection of the study population, data collection, and the association between different determinants of an effect in the population. A definition of the most common biases occurring in these stages is given.

Neural networks and the bias/variance dilemma

Stuart Geman — 2005-07-27T20:12:58-00:00

Neural Comput., Vol. 4, No. 1. (January 1992), pp. 1-58.

Arguments for Randomizing

Patrick Suppes — 2005-03-15T12:29:47-00:00

Three main lines of arguments are presented as a defense of randomization in experimental design. The first concerns the computational advantages of randomizing when a well-defined underlying theoretical model is not available, as is often the case in much experimentation in the medical and social sciences. The high desirability, even for the most dedicated Bayesians, of physical randomization in some special cases is stressed. The second line of argument concerns communication of methodology and results, especially in terms of concerns about bias. The third line of argument concerns the use of randomization to guarantee causal inferences, whether the inference consists of the identification of a prima facie or a genuine cause. In addition, the relation of randomization to measures of complexity and the possibility of accepting only random procedures that produce complex results are discussed.

Is this a question? Not for long. The statement bias

Mario Pandelaere — 2005-09-30T15:04:21-00:00

Journal of Experimental Social Psychology, Vol. In Press, Corrected Proof

Four experiments demonstrate a fundamental 'statement bias': questions are more often misremembered as statements than vice versa. The bias increases with increasing item comprehensibility (Experiment 1) and is related to depth of processing at encoding (Experiment 2). When sentences are simply comprehended, the bias is not affected by the truth of the statement underlying the sentence (Experiment 3). The statement bias generalizes to contexts in which people have to express consent with the content of the sentence (Experiment 4) but is somewhat reduced when they are not sure what the correct answer is. Our findings are consistent with the idea that during processing of a sentence the content of the sentence is represented similar to a statement.

Threat-Related Attentional Bias in Anxious and Nonanxious Individuals: A Meta-Analytic Study

Yair Bar-Haim — 2007-05-10T15:49:29-00:00

Psychological Bulletin, Vol. 133, No. 1. (January 2007), pp. 1-24.

This meta-analysis of 172 studies (N = 2,263 anxious, N = 1,768 nonanxious) examined the boundary conditions of threat-related attentional biases in anxiety. Overall, the results show that the bias is reliably demonstrated with different experimental paradigms and under a variety of experimental conditions, but that it is only an effect size of d = 0.45. Although processes requiring conscious perception of threat contribute to the bias, a significant bias is also observed with stimuli outside awareness. The bias is of comparable magnitude across different types of anxious populations (individuals with different clinical disorders, high-anxious nonclinical individuals, anxious children and adults) and is not observed in nonanxious individuals. Empirical and clinical implications as well as future directions for research are discussed.

Variations on a human universal: Individual differences in positivity offset and negativity bias

Tiffany Ito — 2004-12-29T21:19:32-00:00

Cognition and Emotion, Vol. 19, No. 1., 1.

Teaching Pigeons to Commit Base-Rate Neglect

Edmund Fantino — 2005-09-23T11:47:23-00:00

Psychological Science, Vol. 16, No. 10. (October 2005), pp. 820-825.

Preference-Consistent Evaluation of Information in the Hidden Profile Paradigm: Beyond Group-Level Explanations for the Dominance of Shared Information in Group Decisions, ,

Tobias Greitemeyer — 2007-06-20T18:58:03-00:00

Journal of Personality and Social Psychology, Vol. 84, No. 2. (February 2003), pp. 322-339.

Common explanations for the failure of groups to solve so-called hidden profiles focus on group processes, namely insufficient discussion of unshared information and premature consensus on a suboptimal alternative. As 2 experiments show, even in the absence of such group processes, hidden profiles are hardly ever solved. In Experiment 1, participants first received individual information about a personnel selection task and then read a group discussion protocol containing full information exchange. If the individual information was misleading (hidden profile), most participants failed to detect the correct alternative. In Experiment 2, it was determined that this effect is due to preference-consistent evaluation of information that constitutes an individual-level process mediating the failure of group members to solve hidden profiles.

Bias Reduction of Maximum Likelihood Estimates

David Firth — 2005-02-21T18:09:03-00:00

Biometrika, Vol. 80, No. 1. (1993), pp. 27-38.

It is shown how, in regular parametric problems, the first-order term is removed from the asymptotic bias of maximum likelihood estimates by a suitable modification of the score function. In exponential families with canonical parameterization the effect is to penalize the likelihood by the Jeffreys invariant prior. In binomial logistic models, Poisson log linear models and certain other generalized linear models, the Jeffreys prior penalty function can be imposed in standard regression software using a scheme of iterative adjustments to the data.

Striking nucleotide frequency pattern at the borders of highly conserved vertebrate non-coding sequences.

K Walter — 2005-08-25T15:46:59-00:00

Trends Genet, Vol. 21, No. 8. (August 2005), pp. 436-440.

In a recent study, 1373 highly conserved non-coding elements (CNEs) were detected by aligning the human and Takifugu rubripes (Fugu) genomes. The remarkable degree of sequence conservation in CNEs compared with their surroundings suggested comparing the base composition within CNEs with their 5' and 3' flanking regions. The analysis reveals a novel, sharp and distinct signal of nucleotide frequency bias precisely at the border between CNEs and flanking regions.

Hindsight bias and outcome-consistent thoughts when observing and making service provider decisions

Therese Louie — 2005-08-22T15:35:42-00:00

Organizational Behavior and Human Decision Processes, Vol. 98, No. 1. (September 2005), pp. 88-95.

Two studies examined the relationship between hindsight bias and corresponding open-ended thoughts for decisions in a service provider setting. Perspectives of those observing and making decisions were examined. In study 1, business students who learned the results of a financial advisor's stock purchase showed the traditional hindsight effect regardless of outcome favorability, and produced heightened outcome-consistent thoughts. In contrast, study 2 participants were asked to make (rather than to observe) a purchase decision. They produced hindsight bias and consistent thoughts only when their decision outcomes were favorable. Relative to observers, those acting as the service provider (a) showed more bias when outcomes were favorable, and (b) showed less bias when outcomes were unfavorable. Discussion focuses on hindsight implications in service provider decision-making settings.

Controlling Racial Prejudice: Social-Cognitive Goals Affect Amygdala and Stereotype Activation

Mary Wheeler — 2005-01-07T16:22:43-00:00

Psychological Science, Vol. 16, No. 1., 56.

The malleability of stereotyping matters in social psychology and in society. Previous work indicates rapid amygdala and cognitive responses to racial out-groups, leading some researchers to view these responses as inevitable. In this study, the methods of social-cognitive neuroscience were used to investigate how social goals control prejudiced responses. Participants viewed photographs of unfamiliar Black and White faces, under each of three social goals: social categorization (by age), social individuation (vegetable preference), and simple visual inspection (detecting a dot). One study recorded brain activity in the amygdala using functional magnetic resonance imaging, and another measured cognitive activation of stereotypes by lexical priming. Neither response to photos of the racial out-group was inevitable; instead, both responses depended on perceivers' current social-cognitive goal.

Automatic stereotypes vs. automatic prejudice: Sorting out the possibilities in the weapon paradigm

CM Judd — 2004-12-28T16:58:08-00:00

Journal of Experimental Social Psychology, Vol. 40, No. 1. (January 2004), pp. 75-81.

Payne (2001) has documented that African-American faces automatically facilitate the categorization of handguns, relative to White faces. We suggest that these provocative results could derive from either the automatic activation of prejudice (negative evaluations) or the automatic activation of stereotypes (both positively and negatively valenced associations). In an extension of Payne’s procedure, we show that African-American faces facilitate the categorization of both handguns and sports-related objects, but not the categorization of insects or fruits. Additionally, both handguns and sports objects are more likely to be miscategorized following a White face prime than an African-American one. These results suggest that when perceivers are attempting to identify objects, automatic stereotypic associations, both positively and negatively valenced ones, are more influential than general negative sentiments towards African-Americans.

A Bayesian view of covariation assessment

Craig Mckenzie — 2006-07-24T10:47:18-00:00

Cognitive Psychology, Vol. In Press, Corrected Proof

When participants assess the relationship between two variables, each with levels of presence and absence, the two most robust phenomena are that: (a) observing the joint presence of the variables has the largest impact on judgment and observing joint absence has the smallest impact, and (b) participants' prior beliefs about the variables' relationship influence judgment. Both phenomena represent departures from the traditional normative model (the phi coefficient or related measures) and have therefore been interpreted as systematic errors. However, both phenomena are consistent with a Bayesian approach to the task. From a Bayesian perspective: (a) joint presence is normatively more informative than joint absence if the presence of variables is rarer than their absence, and (b) failing to incorporate prior beliefs is a normative error. Empirical evidence is reported showing that joint absence is seen as more informative than joint presence when it is clear that absence of the variables, rather than their presence, is rare.

The Impact of Emotion on Perception: Bias or Enhanced Processing?

Rene Zeelenberg — 2006-03-22T05:32:00-00:00

Psychological Science, Vol. 17, No. 4. (April 2006), pp. 287-291.

Center-of-inattention: Position biases in decision-making

Priya Raghubir — 2006-03-03T16:34:49-00:00

Organizational Behavior and Human Decision Processes, Vol. 99, No. 1. (January 2006), pp. 66-80.

This paper examines centrality of physical position as a cue that leads to systematic biases in people's decisions to retain or eliminate a participant from a group. Termed the "center-stage" effect, we argue that people use their belief that "important people sit in the middle" as a schematic cue that they substitute for individuating performance information for individuals who occupy central positions when the goal is to eliminate all but one of the group members. This leads to the errors of those in center-positions being overlooked: or making them the "centers-of-inattention." Study 1 examines people's lay beliefs regarding positions using two stylized placement tasks (a group interview and classroom seating scenarios). These suggest that people believe that more attention is paid to those in the center than those on the extremes. Study 2 tests the center-stage effect using observational data from a real television show, The Weakest Link. Results show that players assigned at random to central positions are more likely to win the game than those in extreme positions. Study 3, a laboratory experiment manipulating attention paid to the game shows that observers overlook the errors of players in the center to a greater extent than the errors of players in extreme positions. Study 4 replicates the game in the laboratory with direct process measures to show that players playing the game make the same error. Study 5 shows that in a stylized group interview setting, participants who believe that "important people sit in the middle" find the performance of candidates in the extreme position easier to recall than the performance of those in the central position, and are more likely to choose them. Study 6 shows that the "center-stage" effects are weaker when the end-game rule allows for two (vs one) contestants to be retained. Overall results converge to show that the use of the "center-stage" heuristic substitutes for the effortful processing of individuating information, leading to a biased (favorable) assessment of people in the center. Implications for decision-making are discussed.