A 2 × 2 taxonomy of multilevel latent contextual models: accuracy-bias trade-offs in full and partial error correction models

In multilevel modeling, group-level variables (L2) for assessing contextual effects are frequently generated by aggregating variables from a lower level (L1). A major problem of contextual analyses in the social sciences is that there is no error-free measurement of constructs. In the present article, 2 types of error occurring in multilevel data when estimating contextual effects are distinguished: unreliability that is due to measurement error and unreliability that is due to sampling error. The fact that studies may or may not correct for these 2 types of error can be translated into a 2 × 2 taxonomy of multilevel latent contextual models comprising 4 approaches: an uncorrected approach, partial correction approaches correcting for either measurement or sampling error (but not both), and a full correction approach that adjusts for both sources of error. It is shown mathematically and with simulated data that the uncorrected and partial correction approaches can result in substantially biased estimates of contextual effects, depending on the number of L1 individuals per group, the number of groups, the intraclass correlation, the number of indicators, and the size of the factor loadings. However, the simulation study also shows that partial correction approaches can outperform full correction approaches when the data provide only limited information in terms of the L2 construct (i.e., small number of groups, low intraclass correlation). A real-data application from educational psychology is used to illustrate the different approaches.     

The case of Pinocchio: teachers’ ability to detect deception

In a study with 365 teacher students, 447 teacher trainees, and 123 teachers, the ability to detect students’ deception was tested. Participants judged the credibility of videotaped students who were accused of academic dishonesty (having cheated in a test). Half of these messages were actually true (students had not cheated on the test) and half of them were deceptive (students had cheated on the test). As expected and in line with findings on the influence of expertise on the ability to detect deception from other fields, we found that the overall accuracy rate of teachers was not higher than that of teacher trainees and teacher students. Moreover, we found no effect of teaching experience (years working as a teacher) on overall detection of deception accuracy. Interestingly, teachers were found to have a stronger truth bias and therefore had a lower accuracy in detecting deceptive messages than teacher students and teacher trainees (veracity effect). While teacher characteristics accounted for very little variance, senders’ opportunity to prepare and their gender had strong effects. Detection accuracy was higher for messages where the student had no chance to prepare before being accused of cheating. Overall, independent, or experience, participants hold inaccurate beliefs about deception.     

Short assessment of the Big Five: robust across survey methods except telephone interviewing

We examined measurement invariance and age-related robustness of a short 15-item Big Five Inventory (BFI-S) of personality dimensions, which is well suited for applications in large-scale multidisciplinary surveys. The BFI-S was assessed in three different interviewing conditions: computer-assisted or paper-assisted face-to-face interviewing, computer-assisted telephone interviewing, and a self-administered questionnaire. Randomized probability samples from a large-scale German panel survey and a related probability telephone study were used in order to test method effects on self-report measures of personality characteristics across early, middle, and late adulthood. Exploratory structural equation modeling was used in order to test for measurement invariance of the five-factor model of personality trait domains across different assessment methods. For the short inventory, findings suggest strong robustness of self-report measures of personality dimensions among young and middle-aged adults. In old age, telephone interviewing was associated with greater distortions in reliable personality assessment. It is concluded that the greater mental workload of telephone interviewing limits the reliability of self-report personality assessment. Face-to-face surveys and self-administrated questionnaire completion are clearly better suited than phone surveys when personality traits in age-heterogeneous samples are assessed.     

The Hormonal Correlates of Implicit Motives

Implicit motives represent nonconsciously represented dispositions to seek specific classes of incentives and to avoid corresponding classes of disincentives. Growing evidence suggests that the implicit power motive is associated with basal levels and reactivity of the gonadal steroid hormones testosterone in men and estradiol in women. It is also associated with increased release of stress hormones (cortisol, norepinephrine) in response to dominance challenges and social defeat. The implicit affiliation motive is linked to the release of progesterone: increases in progesterone are followed by increases in affiliation motivation, and arousal of affiliation motivation is associated with concurrent or subsequent progesterone increases. There is limited evidence for a role of vasopressin in achievement motivation. These findings point to a key role of the hypothalamus for implicit motives, a role that is consistent with the existence of function‐specific nuclei within this brain area.     

Implicit theories of emotion: affective and social outcomes across a major life transition

The authors demonstrate that people differ systematically in their implicit theories of emotion: Some view emotions as fixed (entity theorists), whereas others view emotions as more malleable (incremental theorists). Using a longitudinal and multimethod design, the authors show that implicit theories of emotion, as distinct from intelligence, are linked to both emotional and social adjustment during the transition to college. Before entering college, individuals who held entity (vs. incremental) theories of emotion had lower emotion regulation self-efficacy and made less use of cognitive reappraisal (Part 1). Throughout their first academic term, entity theorists of emotion had less favorable emotion experiences and received decreasing social support from their new friends, as evidenced by weekly diaries (Part 2). By the end of freshman year, entity theorists of emotion had lower well-being, greater depressive symptoms, and lower social adjustment as indicated in both self- and peer-reports (Part 3). The emotional, but not the social, outcomes were partially mediated by individual differences in emotion regulation self-efficacy (Part 4). Together, these studies demonstrate that implicit theories of emotion can have important long-term implications for socioemotional functioning.     

Reducing stereotype threat by blurring intergroup boundaries

The authors aimed to establish whether interventions designed to reduce intergroup bias could be applied to the stereotype threat domain. In three experiments, the hypothesis was tested that blurring intergroup boundaries would reduce stereotype threat. In the first study, it was found that female participants who thought about characteristics shared between the genders tended to show less preference for stereotypical female careers than did participants in the baseline condition. In Experiment 2, participants who thought about overlapping characteristics answered more math questions correctly compared to a baseline group and participants who thought about differences between the genders. In Experiment 3, a specific threat manipulation was included. Participants who completed the overlapping characteristics task before receiving the threat completed significantly more math questions correctly than did participants in the baseline and threat conditions. The findings support the idea that interventions designed to reduce intergroup bias can be applied successfully in the reduction of stereotype threat.     

A Third Method of Ethics?

In recent decades, the idea has become common that so‐called virtue ethics constitutes a third option in ethics in addition to consequentialism and deontology. This paper argues that, if we understand ethical theories as accounts of right and wrong action, this is not so. Virtue ethics turns out to be a form of deontology (that is, non‐consequentialism). The paper then moves to consider the Aristotelian distinction between right or virtuous action on the one hand, and acting rightly or virtuously on the other. It is claimed that virtue might play an important role in an explanation of acting virtuously (as it does in Aristotle’s ethics), but that such explanations can be charged with ‘double‐counting’ the moral value of the virtues. The paper concludes that, if we focus on the question of the value of virtue, rather than on the notion of right action, there is room for a self‐standing and important view which could be described as virtue ethics.