Karl Pearson and Eugenics: Personal Opinions and Scientific Rigor

The influence of personal opinions and biases on scientific conclusions is a threat to the advancement of knowledge. Expertise and experience does not render one immune to this temptation. In this work, one of the founding fathers of statistics, Karl Pearson, is used as an illustration of how even the most talented among us can produce misleading results when inferences are made without caution or reference to potential bias and other analysis limitations. A study performed by Pearson on British Jewish schoolchildren is examined in light of ethical and professional statistical practice. The methodology used and inferences made by Pearson and his coauthor are sometimes questionable and offer insight into how Pearson’s support of eugenics and his own British nationalism could have potentially influenced his often careless and far-fetched inferences. A short background into Pearson’s work and beliefs is provided, along with an in-depth examination of the authors’ overall experimental design and statistical practices. In addition, portions of the study regarding intelligence and tuberculosis are discussed in more detail, along with historical reactions to their work.     

Individual Differences In Intra-Individual Variability

For each of 75 Ss several estimates of intra-individual variations were calculated for each of 4 motor skills. The reliability of intra-individual variation for these data, estimated from Pearson productmoment correlations, ranged from r = .31 to r = .51 (in all cases p < .05). It was also found that in 3 of the 4 skills, intra-individual variation correlated significantly (r = .36, .54, .55) with mean performance level. Second order partial correlations, holding constant the effect of mean performance level, were then used to estimate the reliability of intravariation. These correlations were statistically significant and ranged from r = .27 to r = .34. When correlations of intra-individual variation were calculated among the four skills it was demonstrated that intra-individual variation was approximately 95 to 100% task specific.     

The Bias of Crowds: How Implicit Bias Bridges Personal and Systemic Prejudice

As public awareness of implicit bias has grown in recent years, studies have raised important new questions about the nature of implicit bias effects. First, implicit biases are widespread and robust on average, yet are unstable across a few weeks. Second, young children display implicit biases indistinguishable from those of adults, which suggests to many that implicit biases are learned early. Yet, if implicit biases are unstable over weeks, how can they be stable for decades? Third, meta-analyses suggest that individual differences in implicit bias are associated weakly, although significantly, with individual differences in behavioral outcomes. Yet, studies of aggregate levels of implicit bias (i.e., countries, states, counties) are strongly associated with aggregate levels of disparities and discrimination. These puzzles are difficult to reconcile with traditional views, which treat implicit bias as an early-learned attitude that drives discrimination among individuals who are high in bias. We propose an alternative view of implicit bias, rooted in concept accessibility. Concept accessibility can, in principle, vary both chronically and situationally. The empirical evidence, however, suggests that most of the systematic variance in implicit bias is situational. Akin to the “wisdom of crowds” effect, implicit bias may emerge as the aggregate effect of individual fluctuations in concept accessibility that are ephemeral and context-dependent. This bias of crowds theory treats implicit bias tests as measures of situations more than persons. We show how the theory can resolve the puzzles posed and generate new insights into how and why implicit bias propagates inequalities.     

Substantive and operational issues of response bias across levels of analysis: an example of climate-satisfaction relationships

Two studies tested whether method variance is present at multiple levels of analysis and whether methodological procedures can minimize its impact. In Study 1, 8,052 employees from 71 hotels completed measures of climate, work environment characteristics, and satisfaction. A comparison of correlations at the individual level, cross-level, cross-level split, aggregate level, and aggregate-split level of analysis revealed that response bias was present across multiple levels. Results suggest that samples should be split in half when cross-level and aggregate correlations are computed to ameliorate response bias problems that arise from individual-level method variance. In Study 2, results indicated that the temporal spacing of measures of climate and satisfaction influenced response bias. Implications and recommendations for future research are discussed.     

Characterizing belief bias in syllogistic reasoning: A hierarchical Bayesian meta-analysis of ROC data

The belief-bias effect is one of the most-studied biases in reasoning. A recent study of the phenomenon using the signal detection theory (SDT) model called into question all theoretical accounts of belief bias by demonstrating that belief-based differences in the ability to discriminate between valid and invalid syllogisms may be an artifact stemming from the use of inappropriate linear measurement models such as analysis of variance (Dube et al., Psychological Review, 117(3), 831–863, 2010). The discrepancy between Dube et al.’s, Psychological Review, 117(3), 831–863 (2010) results and the previous three decades of work, together with former’s methodological criticisms suggests the need to revisit earlier results, this time collecting confidence-rating responses. Using a hierarchical Bayesian meta-analysis, we reanalyzed a corpus of 22 confidence-rating studies (N =?993). The results indicated that extensive replications using confidence-rating data are unnecessary as the observed receiver operating characteristic functions are not systematically asymmetric. These results were subsequently corroborated by a novel experimental design based on SDT’s generalized area theorem. Although the meta-analysis confirms that believability does not influence discriminability unconditionally, it also confirmed previous results that factors such as individual differences mediate the effect. The main point is that data from previous and future studies can be safely analyzed using appropriate hierarchical methods that do not require confidence ratings. More generally, our results set a new standard for analyzing data and evaluating theories in reasoning. Important methodological and theoretical considerations for future work on belief bias and related domains are discussed.     

Psychological biases of individual investors and financial satisfaction

Traditional finance theory is based on the principle of maximization of utility and explains how choices are made by rational people. Although the theory provides numerous insights, observation of actual behavior of people was seen to be different from what the theory predicted. The homo economicus is in reality a homo sapien who has emotions and beliefs that help to filter the content from his or her environment. These beliefs and preferences that arise due to cognitive limitations, presence of emotions, and various psychological motives guide or bias his or her decisions. Much literature states that the biases should be corrected as they negatively impact financial behaviour and individual's well‐being. However, evolutionary psychology considers biases as design features of human mind. Thus, biases are not always bad, as at times, these biases can help the individual investor to choose the best course of action from the multiple possibilities and enable committing the less costly mistakes, thereby helping the individual to achieve satisficing behaviour. This paper aims to explore the investor biases and see whether they are related to the financial satisfaction of the individuals. Financial satisfaction is the measure of satisfaction with one's financial situation. The results showed that overconfidence bias, reliance on expert bias, and self‐control bias have a positive and significant association with financial satisfaction levels. Association of a few other biases with financial satisfaction was also observed under certain control conditions. This study provides further insights on investor behavior and paves the way for various possibilities for future research.     

Combined behavioural markers of cognitive biases are associated with anhedonia

Biases towards negative information, as well as away from positive information, are associated with psychopathology. Examining biases in multiple processes has been theorised to be more predictive than examining bias in any process alone. Anhedonia is a core symptom of psychopathology and predictive of future psychopathological symptoms. Finding that combined biases are associated with anhedonia would advance knowledge of the nature of emotional processing biases and the value of objective performance-based measures for identifying early risk markers. Participants (N?=?139) completed tasks that assess latency bias (dot probe) and biased recognition (two-alternative forced-choice) of emotional information, as well as an anhedonia measure. An index was computed for each task’s performance reflecting biased processing of positive and negative words. Only combined biases on both tasks were associated with anhedonia. Attentional bias was positively associated with anhedonia, but only when recognition bias for emotional words was high. Thus, assessing biases in multiple domains increased sensitivity to uncover relationships between emotional processing biases and anhedonic symptoms.     

Threat-related attentional bias in anxious and nonanxious individuals: a meta-analytic study

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.     

Second choices in a visual span of apprehension task

Four Ss were run in a visual span of apprehension experiment to determine whether second choices made following incorrect first responses are at the chance level, as implied by various high threshold models proposed for this situation. The relationships between response biases on first and second choices, and between first choice biases on trials with two or three possible responses, were also examined in terms of Luce’s (1959) choice theory. The results were: (a) second choice performance in this task appears to be determined by response bias alone, i.e., second choices were at the chance level; (b)first and second choice response biases were not related according to Luce’s choice axiom; and (c) the choice axiom predicted with reasonable accuracy the relationships between first choice response biases corresponding to trials with different numbers of possible response alternatives.     

Fixed Effects vs. Random Effects Meta-Analysis Models: Implications for Cumulative Research Knowledge

Research conclusions in the social sciences are increasingly based on meta-analysis, making questions of the accuracy of meta-analysis critical to the integrity of the base of cumulative knowledge. Both fixed effects (FE) and random effects (RE) meta-analysis models have been used widely in published meta-analyses. This article shows that FE models typically manifest a substantial Type I bias in significance tests for mean effect sizes and for moderator variables (interactions), while RE models do not. Likewise, FE models, but not RE models, yield confidence intervals for mean effect sizes that are narrower than their nominal width, thereby overstating the degree of precision in meta-analysis findings. This article demonstrates analytically that these biases in FE procedures are large enough to create serious distortions in conclusions about cumulative knowledge in the research literature. We therefore recommend that RE methods routinely be employed in meta-analysis in preference to FE methods.     

The Flexible Correction Model: Bias Correction Guided by Naïve Theories of Bias

Psychological researchers have examined a broad array of biases and shortcomings of social perceivers. Less attention has been paid to how people react when they become concerned about the possibility of bias and attempt to correct or overcome the potential for bias. An early approach was to think of bias correction as “subtracting” information or reactions that are identified as coming from the biasing factor instead of the target. An alternative approach is to conceptualize bias correction as guided by social perceivers' naïve theories or perceptions of the bias at work. The subtraction theories were designed to deal with the assimilative biases typical of early priming experiments, but theory‐based corrections can readily deal with either assimilative or contrastive biases. We describe both types of correction theories, with a focus on the empirical support for theory‐based correction (and especially for the Flexible Correction Model). We also identify future directions for research on theory‐based correction.     

Weight Bias Internalization and Psychosocial,Physical, and Behavioral Health: A Meta-Analysis of Cross-Sectional and Prospective Associations

Coinciding with widespread efforts to address obesity, weight bias internalization (a process of self-devaluation wherein individuals apply weight-biased stereotypes to themselves) has gained increased attention as a robust correlate of poor health outcomes. The present meta-analysis aimed to provide the largest quantitative synthesis of associations between weight bias internalization and health-related correlates. Studies that provided zero-order correlations for cross-sectional or prospective associations between weight bias internalization and physical, psychosocial, and behavioral health correlates were included in the meta-analysis. Meta-regression determined whether these associations differed based on demographic (sex/gender, race, age), anthropometric (body mass index), and study-level (publication status, sample type, study quality) moderators. Data for 149 (sub)samples were identified that included between 14 and 18,766 participants (M sample size = 534.96, SD = 1,914.43; M age = 34.73, SD = 12.61, range = 9.95–65.70). Results indicated that greater weight bias internalization was concurrently associated with worse psychosocial (e.g., negative and positive mental health, social functioning), physical (e.g., BMI, weight maintenance, health-related quality of life [HRQoL]), and behavioral health (e.g., disordered eating behaviors, healthy eating, physical activity) across most constructs, with effects ranging from small to very large in magnitude. Preliminary evidence also suggested that greater weight bias internalization was subsequently associated with less weight loss and increased negative mental health. Notable variations in the nature and magnitude of these associations were identified based on the health-related correlate and moderator under consideration. These findings indicate that weight bias internalization is linked to multiple adverse health-related outcomes and provide insight into priorities for future research, theory building, and interventions in this area.     

Using full matching to estimate causal effects in nonexperimental studies: examining the relationship between adolescent marijuana use and adult outcomes

Matching methods such as nearest neighbor propensity score matching are increasingly popular techniques for controlling confounding in nonexperimental studies. However, simple k:1 matching methods, which select k well-matched comparison individuals for each treated individual, are sometimes criticized for being overly restrictive and discarding data (the unmatched comparison individuals). The authors illustrate the use of a more flexible method called full matching. Full matching makes use of all individuals in the data by forming a series of matched sets in which each set has either 1 treated individual and multiple comparison individuals or 1 comparison individual and multiple treated individuals. Full matching has been shown to be particularly effective at reducing bias due to observed confounding variables. The authors illustrate this approach using data from the Woodlawn Study, examining the relationship between adolescent marijuana use and adult outcomes.     

Making heads or tails of probability: An experiment with random generators

Background. The equiprobability bias is a tendency for individuals to think of probabilistic events as ‘equiprobable’ by nature, and to judge outcomes that occur with different probabilities as equally likely. The equiprobability bias has been repeatedly found to be related to formal education in statistics, and it is claimed to be based on a misunderstanding of the concept of randomness. Aims. The aim of the present study was to examine whether experimenting with random generators would decrease the equiprobability bias. Sample. The participants were 108 psychology students whose performance was measured either immediately after taking part in a training session (n= 55), or without doing any training exercises (n= 53). Method. The training session consisted of four activities. These included generating random sequences of events, and learning about the law of large numbers. Subsequently, the participants were tested on a series of equiprobability problems, and a number of other problems with similar structure and content. Results. The results indicated that the training successfully decreased the equiprobability bias. However, this effect was moderated by participants’ cognitive ability (i.e., higher ability participants benefitted from the training more than participants with lower cognitive ability). Finally, the training session had the unexpected side effect of increasing students’ susceptibility to the representativeness heuristic. Conclusions. Experimenting with random generators has a positive effect on students’ general understanding of probability, but the same time it might increase their susceptibility to certain biases (especially, to the representativeness heuristic). These findings have important implications for using training methods to improve probabilistic reasoning performance.     

Examining Social Desirability in Measures of Religion and Spirituality Using the Bogus Pipeline

A primary concern in the psychology of religion is the distinct possibility that responses to empirical assessments of individuals’ degree and type of religiosity and spirituality are exaggerated owing to social desirability bias. In spite of increased secularization in American culture and a growing distrust of organized religion, religious involvement, personal religiosity, and spirituality are still viewed as highly desirable characteristics. This study estimates the extent of social desirability biases that affect self-reports of religion and spirituality by utilizing a bogus pipeline procedure. In this procedure, participants are convinced that experimenters can detect disingenuous responses to individual items on questionnaires through the use of physiological measures, although no physiological data are actually collected. If the self-reports of participants in the bogus pipeline condition indicate greater religiosity or spirituality than those in the control condition, self-report bias is indicated. The bogus pipeline procedure has been used in other areas of study to increase veracity of self-reports when social desirability effects are present (such as reporting sexual behaviors or prejudice). The results indicate that social desirability biases influence multiple constructs including religious orientations, religious coping, and daily spiritual experiences. Implications for future research relying on self-reports of religion and spirituality are discussed.     

Buyers’ Trust of the Salesperson: An Item-Level Meta-Analysis

The issue of trust is an important one in the marketing literature. To assess the nomological validity of trust-related measures, this research uses an item-level measurement meta-analysis based on correlations from 32 studies that measure trust-related constructs in the context of a buyer’s assessments of sellers. Analysis reveals that 16-item measures converge into three constructs that are indicative of the seller’s credibility, expertise, and compatibility and three items each that are indicative of trust and trustworthiness. In a structural equation model based on the meta-analysis, a seller’s characteristics are predictive of trust and trustworthiness.     

Attention biases and habituation of attention biases are associated with 5-HTTLPR and COMTval158met

Fundamental biases in affective information processing are modulated by individual differences in the emotional response to environmental stimuli that may be partly based on the individual’s genetic make-up. To extend prior dot probe studies on attention genetics, we used a visual-search paradigm (VSP) with pictures of angry and happy faces of both sexes as targets, neutral faces as distractors, and a varying set size. Participants were selected a priori depending on their 5-HTTLPR (s/s, s/l, l/l; on a constant rs25531 A-allele background) and COMTval158met (val/val, valmet, met/met) genotypes and were matched for sex and age. We demonstrate a bias towards angry male faces (as opposed to happy male faces) irrespective of 5-HTTLPR genotype in the first experimental block that was maintained during the second experimental block only in carriers of the s-allele, which implies differential habituation processes. While a bias towards angry male faces was observed irrespective of COMTval158met genotype, only individuals with the val/val genotype exhibited a bias towards a happy female face (as opposed to an angry female face). In sum, our results both replicate and extend prior findings in the field of attention genetics and add important pieces of information to the research on attentional biases in emotion processing.     

A meta-analysis of the effect of cognitive bias modification on anxiety and depression

Cognitive biases have been theorized to play a critical role in the onset and maintenance of anxiety and depression. Cognitive bias modification (CBM), an experimental paradigm that uses training to induce maladaptive or adaptive cognitive biases, was developed to test these causal models. Although CBM has generated considerable interest in the past decade, both as an experimental paradigm and as a form of treatment, there have been no quantitative reviews of the effect of CBM on anxiety and depression. This meta-analysis of 45 studies (2,591 participants) assessed the effect of CBM on cognitive biases and on anxiety and depression. CBM had a medium effect on biases (g = 0.49) that was stronger for interpretation (g = 0.81) than for attention (g = 0.29) biases. CBM further had a small effect on anxiety and depression (g = 0.13), although this effect was reliable only when symptoms were assessed after participants experienced a stressor (g = 0.23). When anxiety and depression were examined separately, CBM significantly modified anxiety but not depression. There was a nonsignificant trend toward a larger effect for studies including multiple training sessions. These findings are broadly consistent with cognitive theories of anxiety and depression that propose an interactive effect of cognitive biases and stressors on these symptoms. However, the small effect sizes observed here suggest that this effect may be more modest than previously believed.     

Young macaques (<Emphasis Type="Italic">Macaca fascicularis</Emphasis>) preferentially bias attention towards closer,older, and better tool users

Examining how animals direct social learning during skill acquisition under natural conditions, generates data for examining hypotheses regarding how transmission biases influence cultural change in animal populations. We studied a population of macaques on Koram Island, Thailand, and examined model-based biases during interactions by unskilled individuals with tool-using group members. We first compared the prevalence of interactions (watching, obtaining food, object exploration) and proximity to tool users during interactions, in developing individuals (infants, juveniles) versus mature non-learners (adolescents, adults), to provide evidence that developing individuals are actively seeking information about tool use from social partners. All infants and juveniles, but only 49% of mature individuals carried out interacted with tool users. Macaques predominantly obtained food by scrounging or stealing, suggesting maximizing scrounging opportunities motivates interactions with tool users. However, while interactions by adults was limited to obtaining food, young macaques and particularly infants also watched tool users and explored objects, indicating additional interest in tool use itself. We then ran matrix correlations to identify interaction biases, and what attributes of tool users influenced these. Biases correlated with social affiliation, but macaques also preferentially targeted tool users that potentially increase scrounging and learning opportunities. Results suggest that social structure may constrain social learning, but the motivation to bias interactions towards tool users to maximize feeding opportunities may also socially modulate learning by facilitating close proximity to better tool users, and further interest in tool-use actions and materials, especially during development.     

The relationship between anti-gay prejudice and the categorization of sexual orientation

A relatively large literature has demonstrated that sexual orientation can be judged accurately from a variety of minimal cues, including facial appearance. Untested in this work, however, is the influence that individual differences in prejudice against gays and lesbians may exert upon perceivers’ judgments. Here, we report the results of a meta-analysis of 23 unpublished studies testing the relationship between anti-gay bias and the categorization of sexual orientation from faces. Aggregating data from multiple measures of bias using a variety of methods in three different countries over a period of 8 years, we found a small but significant negative relationship between accuracy and prejudice that was homogeneous across the samples tested. Thus, individuals reporting higher levels of anti-gay bias appear to be less accurate judges of sexual orientation.