Multilevel and Latent Variable Modeling with Composite Links and Exploded Likelihoods

Composite links and exploded likelihoods are powerful yet simple tools for specifying a wide range of latent variable models. Applications considered include survival or duration models, models for rankings, small area estimation with census information, models for ordinal responses, item response models with guessing, randomized response models, unfolding models, latent class models with random effects, multilevel latent class models, models with log-normal latent variables, and zero-inflated Poisson models with random effects. Some of the ideas are illustrated by estimating an unfolding model for attitudes to female work participation. We wish to thank The Research Council of Norway for a grant supporting our collaboration.     

Multilevel Modeling with Correlated Effects

When there exist omitted effects, measurement error, and/or simultaneity in multilevel models, explanatory variables may be correlated with random components, and standard estimation methods do not provide consistent estimates of model parameters. This paper introduces estimators that are consistent under such conditions. By employing generalized method of moments (GMM) estimation techniques in multilevel modeling, the authors present a series of estimators along a robust to efficient continuum. This continuum depends on the assumptions that the analyst makes regarding the extent of the correlated effects. It is shown that the GMM approach provides an overarching framework that encompasses well-known estimators such as fixed and random effects estimators and also provides more options. These GMM estimators can be expressed as instrumental variable (IV) estimators which enhances their interpretability. Moreover, by exploiting the hierarchical structure of the data, the current technique does not require additional variables unlike traditional IV methods. Further, statistical tests are developed to compare the different estimators. A simulation study examines the finite sample properties of the estimators and tests and confirms the theoretical order of the estimators with respect to their robustness and efficiency. It further shows that not only are regression coefficients biased, but variance components may be severely underestimated in the presence of correlated effects. Empirical standard errors are employed as they are less sensitive to correlated effects when compared to model-based standard errors. An example using student achievement data shows that GMM estimators can be effectively used in a search for the most efficient among unbiased estimators. This research was supported by the National Academy of Education/Spencer Foundation and the National Science Foundation, grant number SES-0436274. We thank the editor, associate editor, and referees for detailed feedback that helped improve the paper.     

Combined Medication and CBT for Generalized Anxiety Disorder With African American Participants: Reliability and Validity of Assessments and Preliminary Outcomes

Using data from a study of combined cognitive behavioral therapy (CBT) and venlafaxine XR in the treatment of generalized anxiety disorder (GAD), the current article examines the reliability and convergent validity of scales, and preliminary outcomes, for African American compared with European American patients. Internal consistency and short-term stability coefficients for African Americans (n = 42) were adequate and similar or higher compared with those found for European Americans (n = 164) for standard scales used in GAD treatment research. Correlations among outcome measures among African Americans were in general not significantly different for African Americans compared with European Americans. A subset of patients with DSM-IV–diagnosed GAD (n = 24 African Americans; n = 52 European Americans) were randomly selected to be offered the option of adding 12 sessions of CBT to venlafaxine XR treatment. Of those offered CBT, 33.3% (n = 8) of the African Americans and 32.6% (n = 17) of the European Americans accepted and attended at least one CBT treatment session. The outcomes for African Americans receiving combined treatment were not significantly different from European Americans receiving combined treatment on primary or secondary efficacy measures.     

Does a Nation's Religious Composition Affect Generalized Trust? The Role of Religious Heterogeneity and the Percent Religious

Is religion more of an integrative or a divisive force in contemporary societies? We use multilevel analyses of World Values Survey data from 77,409 individuals in 69 countries to examine how both the percent of the population that is religious and the religious heterogeneity of a country are related to generalized social trust, the willingness of individuals to trust “most people.” When we first examine the main effects of the percent religious and religious heterogeneity we find no evidence that either variable is related to trust in the ways predicted by major theories. However, the combination of these two variables has a huge negative relationship with trust. Countries that are both highly religious and religiously heterogeneous (diverse) have, on average, levels of trust that are only half the average levels of countries with other combinations of these two variables. The results have important implications for understanding the role of religion in modern societies.     

The logistics of choice

The generalized matching law (GML) is reconstructed as a logistic regression equation that privileges no particular value of the sensitivity parameter, a. That value will often approach 1 due to the feedback that drives switching that is intrinsic to most concurrent schedules. A model of that feedback reproduced some features of concurrent data. The GML is a law only in the strained sense that any equation that maps data is a law. The machine under the hood of matching is in all likelihood the very law that was displaced by the Matching Law. It is now time to return the Law of Effect to centrality in our science.     

Cognitive Behavior Therapy for Women With Generalized Anxiety Disorder in the Perinatal Period: Impact on Problematic Behaviors

Generalized anxiety disorder (GAD) is the most frequently diagnosed anxiety disorder among women in the perinatal period (pregnancy to one year postpartum). Recent studies have examined the relationship between problematic behaviors and GAD symptoms. Studies in nonperinatal samples indicate that adults with GAD engage in avoidance and safety behaviors and these behaviors are associated with greater symptom severity. Little research has examined the use of problematic behaviors among pregnant or postpartum women. However, preliminary research suggests that these behaviors may have a negative impact on both anxious women and their children. Our aim was to examine the extent to which women with GAD in pregnancy or the postpartum engage in problematic behaviors and whether cognitive behavioral therapy is effective in reducing these behaviors. Fifty-eight women with GAD in pregnancy or postpartum were recruited from a larger clinical trial (Clinicaltrials.gov ID NCT02850523) evaluating the effectiveness of group-based cognitive behavioral therapy (CBGT) for perinatal anxiety disorders. The results indicated that women with perinatal GAD reported high levels of avoidance and safety behaviors and greater engagement in these behaviors was associated with higher levels of worry and related symptoms. CBGT was effective in reducing GAD symptoms and problematic behaviors and a bidirectional relationship was found between changes in worry and problematic behaviors during treatment. Limitations and future directions are discussed.     

Random effects and extended generalized partial credit models

In this paper it is shown that under the random effects generalized partial credit model for the measurement of a single latent variable by a set of polytomously scored items, the joint marginal probability distribution of the item scores has a closed-form expression in terms of item category location parameters, parameters that characterize the distribution of the latent variable in the subpopulation of examinees with a zero score on all items, and item-scaling parameters. Due to this closed-form expression, all parameters of the random effects generalized partial credit model can be estimated using marginal maximum likelihood estimation without assuming a particular distribution of the latent variable in the population of examinees and without using numerical integration. Also due to this closed-form expression, new special cases of the random effects generalized partial credit model can be identified. In addition to these new special cases, a slightly more general model than the random effects generalized partial credit model is presented. This slightly more general model is called the extended generalized partial credit model. Attention is paid to maximum likelihood estimation of the parameters of the extended generalized partial credit model and to assessing the goodness of fit of the model using generalized likelihood ratio tests. Attention is also paid to person parameter estimation under the random effects generalized partial credit model. It is shown that expected a posteriori estimates can be obtained for all possible score patterns. A simulation study is carried out to show the usefulness of the proposed models compared to the standard models that assume normality of the latent variable in the population of examinees. In an empirical example, some of the procedures proposed are demonstrated.     

The stability and causal effects of task‐specific and generalized self‐efficacy in college1

This research investigated the stability and causal effects of task‐specific self‐efficacy and generalized self‐efficacy in college students. In Study 1, task‐specific self‐efficacy and generalized self‐efficacy scales were administered to university students (N = 237) on two occasions, with an interval of 2 months. In Study 2, task‐specific (care‐work) self‐efficacy and generalized self‐efficacy scales were administered to college students who were studying to be care workers (N = 49) on three occasions (before the first care‐work practicum, after the practicum, and 3 months later). The results of both studies indicated that generalized self‐efficacy was more stable than task‐specific self‐efficacy. In Study 1, both generalized self‐efficacy and task‐specific self‐efficacy affected each other in daily life. Study 2 examined the effects of an ego‐engaged experience (the first care‐work practicum) for task‐specific (care‐work) self‐efficacy and generalized self‐efficacy. The results indicated that changes in care‐work self‐efficacy did not affect generalized self‐efficacy.