A Bayesian view of covariation assessment

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.     

A Common Genetic Factor Explains the Covariation Among ADHD ODD and CD Symptoms in 9–10 Year Old Boys and Girls

Previous studies examining the covariation among Attention Deficit Hyperactivity Disorder (ADHD), Oppositional Defiant Disorder (ODD) and Conduct Disorder (CD) have yielded inconsistent results. Some studies have concluded that the covariation among these symptoms is due to common genetic influences, whereas others have found a common environmental overlap. The present study investigated the genetic and environmental correlations among these three childhood disorders, based on a sample of 1,219 twins, age 9–10 years. A latent externalizing behavior factor was found to explain the covariance among ADHD, ODD and CD symptoms. Genetic influences explained more than half of the variance in this externalizing factor in both boys and girls. There were also unique genetic and environmental influences in each set of symptoms, suggesting some etiological independence of the three disorders. Our findings have implications for molecular genetic studies trying to identify susceptibility genes for these disorders. This study was funded by NIMH (R01 MH58354). Catherine Tuvblad was supported by post-doctoral stipends from the Swedish Council for Working Life and Social Research (Project 2006-1501) and the Sweden-America Foundation. Adrian Raine was supported by NIMH (Independent Scientist Award K02 MH01114-08). We thank the Southern California Twin Project staff for their assistance in collecting data, and the twins and their families for their participation.     

Does scientific thinking lead to success and sanity? An integration of attribution and attributional models

A motivational model is introduced that integrates conceptions of attributional antecedents with those of attributional consequences, and suggests that it is functional to make realistic attributions. This model is applied to the analysis of achievement behaviour and depression. The use of a new experimental paradigm designed to test the model's assumptions demonstrates that realistic attributions maximise performance. The model was also used to investigate whether depressives are more realistic than non-depressives. This research revealed that depressives draw more "pessimistic" conclusions from covariation information than non-depressives. In addition, depressives expect their partners to make more negative attributions about the depressives' outcomes than their partners in fact do. However, these negative views could not be traced back to faulty inferences from covariation information. Finally, research is described showing that, in addition to depressogenic attributions, certain evaluative cognitions need to be present in order to trigger depressive mood reactions.     

Seeing ghosts: Negative body evaluation predicts overestimation of negative social feedback

The current study investigated whether negative body evaluation predicts women's overestimation of negative social feedback related to their own body (i.e., covariation bias). Sixty-five female university students completed a computer task where photos of their own body, of a control woman's body, and of a neutral object, were followed by nonverbal social feedback (i.e., facial crowds with equal numbers of negative, positive, and neutral faces). Afterward, women estimated the percentage of negative, positive, and neutral social feedback that followed their own body, the control woman's body, and the neutral object. The findings provided evidence for a covariation bias: negative body evaluation predicted higher estimates of negative social feedback for women's own body, but not for the other stimuli. Additionally, the covariation bias was not explained by differences in how women interpreted the social feedback (the facial stimuli). Clinical implications of the covariation bias to body image are discussed.     

Fear-relevant selective associations and social anxiety: absence of a positive bias

An illusory correlation paradigm was used to compare high and low socially anxious individuals' initial, on-line and a posteriori covariation estimates between emotional faces and aversive, pleasant and neutral outcomes. Overall, participants demonstrated an initial expectancy bias for aversive outcomes following angry faces, and pleasant outcomes following happy faces. On-line expectancy biases indicated that initial biases were extinguished during the task, with the exception of low socially anxious individuals who continued to over-associate positive social cues with pleasant outcomes. In addition to lacking this protective positive on-line bias, the high social anxiety group reported retrospectively more negative social cues than the low socially anxious group. Findings are discussed in relation to similar evidence from recent interpretive and memory paradigms.     

Covariation assessment in rank order data

Two experiments were conducted to investigate how individuals assess covariation with rank order data. In both studies, subjects were given sets of rank order data, each set consisting of ten items ranked on two characteristics, and were asked to estimate the degree of relationship for each set. Contrary to previous research, subjects' estimates of covariation in this task were quite sensitive to actual levels of correlation in the data and remained unaffected by simple variations in the way rank order data were presented. More importantly, it appeared that this sensitivity to covariation was due likely to the use of a simple heuristic referred to here as the total discrepancy heuristic. These findings are discussed in terms of the availability of simple heuristics in rank-ordered versus other types of data and the consequences of using such heuristics in decision-making contexts.     

Contamination vs. harm-relevant outcome expectancies and covariation bias in spider phobia

There is increasing evidence that spiders are not feared because of harmful outcome expectancies but because of disgust and contamination-relevant outcome expectancies. This study investigated the relative strength of contamination- and harm-relevant UCS expectancies and covariation bias in spider phobia. High (n=25) and low (n=24) spider fearful individuals saw a series of slides comprising spiders, pitbulls, maggots, and rabbits. Slides were randomly paired with either a harm-relevant outcome (electrical shock), a contamination-related outcome (drinking of a distasting fluid), or nothing. Spider fearful individuals displayed a contamination-relevant UCS expectancy bias associated with spiders, whereas controls displayed a harm-relevant expectancy bias. There was no evidence for a (differential) postexperimental covariation bias; thus the biased expectancies were not robust against refutation. The present findings add to the evidence that contamination ideation is critically involved in spider phobia.     

Role of attention, memory, and covariation-detection processes in clinically significant eating-disorder symptoms

Cognitive theorists implicate a role for cognitive processing of shape- and weight-related information in the maintenance and potential etiology of eating disorders. The present study examines the role of women’s processing of information regarding other women’s body size and affect in eating disorder symptoms among young women. 253 female undergraduates completed attention, memory, and covariation-detection tasks that presented full-body photographs of young women, as well as an eating-disorder measure. High-Symptom women, relative to Medium- and Low-Symptom women, showed greater relative attention to body size than affect, better memory for body size and worse memory for affect, and enhanced sensitivity to the manipulated covariation between body size and affect. All participants perceived an illusory correlation between body size and affect. These findings suggest that cognitive theories may be extended usefully by considering processing of other-relevant and affective information, as well as the role of covariation-detection processes in eating disorders.     

Creating distinct implicit and explicit attitudes with an illusory correlation paradigm

Two studies used an illusory correlation procedure to test whether distinct implicit and explicit evaluations can result from the same learning episode. All participants learned twice as much about the qualities of one group (majority) than another (minority). In one condition, the ratio of positive to negative information was equal between groups. In other conditions, the majority group showed proportionally more positive qualities than the minority group, or vice versa. Participants in the pro-majority and pro-minority conditions formed both implicit and explicit attitudes consistent with the attitude induction. Participants in the illusory correlation condition showed the expected preference for the majority group (the illusory bias), but showed no implicit preference, suggesting distinct influences on implicit and explicit attitude formation. The effects are consistent with dual-process models in which implicit attitudes reflect accounting of covariation and explicit attitudes reflect interpretative judgments of that covariation.     

The association between aggressive and non-aggressive antisocial problems as measured with the Achenbach System of Empirically Based Assessment: A study of 27,861 parent–adolescent dyads from 25 societies

Aggression (e.g., assaulting others, bullying, oppositionality; AGG) and non-aggressive rule-breaking (e.g., lying, stealing, vandalism; RB) appear to constitute meaningfully distinct dimensions of antisocial behavior. Despite these differences, it is equally clear that AGG and RB are moderately-to-strongly intercorrelated with one another. To date, however, we have little insight into the sampling and methodologic characteristics that might moderate the association between AGG and RB. The current study sought to evaluate several such moderators (i.e., age, sex, informant, and society) in a sample of 27,861 parent–adolescent dyads from 25 societies. AGG and RB were assessed with the well-known Child Behavior Checklist and Youth Self-Report (Achenbach & Rescorla, 2001). Results revealed small effects of informant and adolescent sex, such that the association between AGG and RB was stronger for parents’ reports than for adolescents’ self-reports, and for boys than for girls. The association also varied by society. Unexpectedly, the specific operationalization of ‘aggression’ emerged as a particularly strong moderator, such that the association was stronger for a general measure of AGG than for a more focused measure of physical aggression per se. Such findings inform our understanding of similarities and differences between aggressive and non-aggressive antisocial problems.