Using Marginal Structural Modeling for Grade Retention Effects

Vandecandelaere, Vansteelandt, De Fraine, and Van Damme (this issue) described marginal structural modeling (MSM) and used it to estimate the effects of a time-varying intervention, retention (holding back) in school grades, on students' math achievement. This commentary supplements Vandecandelaere et al. (this issue) and discusses several topics in retention studies and MSM. First, we discuss the importance of equating time-varying confounders in retention studies. Second, we discuss same-grade and same-age comparisons in retention studies. Third, we discuss one important section in the authors' overview of MSM: why standard methods (e.g., ANCOVA, propensity score analysis) cannot properly adjust for time-varying confounders. Finally, using the grade retention analyses in Vandecandelaere et al. (this issue) as an example, we provide our insights on four aspects of MSM: (a) covariate selection, (b) estimation of weights, (c) evaluation of balance properties, and (d) missing data handling.     

Causal competition based on generic priors

Although we live in a complex and multi-causal world, learners often lack sufficient data and/or cognitive resources to acquire a fully veridical causal model. The general goal of making precise predictions with energy-efficient representations suggests a generic prior favoring causal models that include a relatively small number of strong causes. Such “sparse and strong” priors make it possible to quickly identify the most potent individual causes, relegating weaker causes to secondary status or eliminating them from consideration altogether. Sparse-and-strong priors predict that competition will be observed between candidate causes of the same polarity (i.e., generative or else preventive) even if they occur independently. For instance, the strength of a moderately strong cause should be underestimated when an uncorrelated strong cause also occurs in the general learning environment, relative to when a weaker cause also occurs. We report three experiments investigating whether independently-occurring causes (either generative or preventive) compete when people make judgments of causal strength. Cue competition was indeed observed for both generative and preventive causes. The data were used to assess alternative computational models of human learning in complex multi-causal situations.     

Le rôle de la stigmatisation dans la symptomatologie dépressive de femmes françaises en surpoids ou en situation d’obésité

IntroductionDespite the high prevalence of depressive symptoms among women being overweight or having obesity, the role of stigma and causal attributions in the expression of these symptoms in French women has been largely under-explored to date.ObjectiveThe aim of this research is to study the role of internalized stigmatization, stigmatizing experiences and causal attribution in the intensity of depressive symptoms in overweight and obese French women.MethodThese four variables were measured in 160 French women being overweight or having obesity.ResultsOur analyses highlight positive correlations between stigmatizing experiences, internalized stigmatization and depressive symptomatology. It is important to note that stigmatizing experiences seem to play a more preponderant role than internalized stigmatization in the expression of depressive symptoms.ConclusionFor women with significant depressive symptoms, it appears essential to integrate interventions aimed at reducing their level of stigma. Moreover, from a preventive perspective, it also seems essential to promote campaigns in France aimed at reducing the stigmatization in this population.     

The Accuracy of Causal Learning Over Long Timeframes: An Ecological Momentary Experiment Approach

The ability to learn cause–effect relations from experience is critical for humans to behave adaptively — to choose causes that bring about desired effects. However, traditional experiments on experience-based learning involve events that are artificially compressed in time so that all learning occurs over the course of minutes. These paradigms therefore exclusively rely upon working memory. In contrast, in real-world situations we need to be able to learn cause–effect relations over days and weeks, which necessitates long-term memory. 413 participants completed a smartphone study, which compared learning a cause–effect relation one trial per day for 24 days versus the traditional paradigm of 24 trials back- to- back. Surprisingly, we found few differences between the short versus long timeframes. Subjects were able to accurately detect generative and preventive causal relations, and they exhibited illusory correlations in both the short and long timeframe tasks. These results provide initial evidence that experience-based learning over long timeframes exhibits similar strengths and weaknesses as in short timeframes. However, learning over long timeframes may become more impaired with more complex tasks.     

Gaussian variational estimation for multidimensional item response theory

Multidimensional item response theory (MIRT) is widely used in assessment and evaluation of educational and psychological tests. It models the individual response patterns by specifying a functional relationship between individuals' multiple latent traits and their responses to test items. One major challenge in parameter estimation in MIRT is that the likelihood involves intractable multidimensional integrals due to the latent variable structure. Various methods have been proposed that involve either direct numerical approximations to the integrals or Monte Carlo simulations. However, these methods are known to be computationally demanding in high dimensions and rely on sampling data points from a posterior distribution. We propose a new Gaussian variational expectation--maximization (GVEM) algorithm which adopts variational inference to approximate the intractable marginal likelihood by a computationally feasible lower bound. In addition, the proposed algorithm can be applied to assess the dimensionality of the latent traits in an exploratory analysis. Simulation studies are conducted to demonstrate the computational efficiency and estimation precision of the new GVEM algorithm compared to the popular alternative Metropolis–Hastings Robbins–Monro algorithm. In addition, theoretical results are presented to establish the consistency of the estimator from the new GVEM algorithm.     

Not Simply in the Eye of the Beholder: Authenticity as a Product of Candidate Preference and Unfettered Speech

Authenticity has emerged over recent decades as a prominent theme in both the press and in political research—and peaked in the 2016 presidential contest that pitted Donald Trump against Hillary Clinton, Bernie Sanders, Marco Rubio, and Ted Cruz. In this context, we attempted to answer the question: How do voters judge a presidential candidate's authenticity? Here we use motivated reasoning and correspondent inference theory as theoretical frameworks to examine how partisan preference combines with perceptions of unfettered speech and strategic impression management to influence voter judgments of a candidate's authenticity. An online survey of 525 respondents demonstrated that individuals' partisan preferences influenced both judgments of a candidate's authenticity and their perceptions of behaviors signifying authenticity (use of unfettered speech versus strategic impression management). These behavioral signals partially mediated the relation between candidate preferences and authenticity judgments. Moreover, voters, given their partisan preferences, differentially weighted candidates' use of unfettered speech and strategic impression management tactics in their judgments of authenticity. Finally, unfiltered/politically incorrect speech was found to have both positive and negative effects on authenticity judgments. Findings further elucidate the nature of authenticity as perceived in others and identify intermediary variables and boundary conditions that influence those perceptions.     

Naïve and Robust: Class‐Conditional Independence in Human Classification Learning

Humans excel in categorization. Yet from a computational standpoint, learning a novel probabilistic classification task involves severe computational challenges. The present paper investigates one way to address these challenges: assuming class‐conditional independence of features. This feature independence assumption simplifies the inference problem, allows for informed inferences about novel feature combinations, and performs robustly across different statistical environments. We designed a new Bayesian classification learning model (the dependence‐independence structure and category learning model, DISC‐LM) that incorporates varying degrees of prior belief in class‐conditional independence, learns whether or not independence holds, and adapts its behavior accordingly. Theoretical results from two simulation studies demonstrate that classification behavior can appear to start simple, yet adapt effectively to unexpected task structures. Two experiments—designed using optimal experimental design principles—were conducted with human learners. Classification decisions of the majority of participants were best accounted for by a version of the model with very high initial prior belief in class‐conditional independence, before adapting to the true environmental structure. Class‐conditional independence may be a strong and useful default assumption in category learning tasks.     

Of BOLD Claims and Excessive Fears: A Call for Caution and Patience Regarding Political Neuroscience

Published and informal assessments of the prospects for neuroimaging in political science have tended to range from overexuberant to reflexively dismissive. We seek to present a cautious but fair middle ground in considering this new methodology, primarily from an epistemological perspective. Our examination centers on the relationship between two levels of analysis, focusing on the potential for connection between behavior‐based theories of political psychology and cognition and the neural processes and systems involved in generating behaviors and states of mind. We explore the place of each level of analysis on its own, as well as the potential for the fruitful interaction of the two. This analysis brings together opinions and ideas presented by others in various forums and across multiple disciplines, offers a discussion of the the promises and perils of neuroimaging in its application to social science, as well as some practical thoughts regarding its early‐stage incorporation into political psychology. We argue in favor of proceeding with more substantial incorporation of neuroimaging into political psychology's methodological arsenal, but note that this will initially require both (1) greater acceptance of work more focused on presenting empirical results than on providing dispositive evidence in broader theoretical debates and (2) a commitment on the part of those conducting this research to refrain from overstating the definitiveness of its theoretical implications.     

Cheapened altruism: Discounting personally affected prosocial actors

Are charitable donors always perceived as charitable? Three studies suggest that although having a personal connection to a cause motivates much charitable giving, donors who have been personally affected by the target cause are given less “credit” for their donations, i.e., are perceived as less intrinsically charitable. These donors are perceived as having selfish motivations even when they have nothing economic or social to gain from the donation. More specifically, personally-affected donors are perceived as driven by emotional selfishness, or a desire to improve their own hedonic state rather a desire to improve the welfare of others, which lessens the charitable credit that they receive. In addition, although donors who have been personally affected by the target cause are seen as less charitable, they are perceived more favorably in other ways (e.g., more loyal).