Search

Recent developments in the collection and modeling of intensive longitudinal data have enabled us to fit dynamic twin models, in which within-person processes are separated into genetic and environmental components. A well-known dynamic twin model is the genetic simplex model, which is fitted to a few repeated measures for many twins. A more recently developed model is the iFACE model, which is fitted to many repeated measures for a single pair of twins. In this paper, we introduce a missing link between these two models–a multilevel extension that allows for making both population-level and twin-level inferences. We provide a proof-of-principle simulation study for this model, and apply it to an experience sampling data set on 148 monozygotic and 88 dizygotic twins. We use the multilevel model to examine the overlap and differences between the dynamic genetic twin models and the classic twin models, as well as their interpretation.

Contains fulltext : 284223.pdf (Publisher’s version ) (Open Access) Background: The adaptive ability performance test (ADAPT) was developed to assess adaptive skills in individuals with intellectual disabilities and borderline intellectual functioning, with or without mental disorders. As a follow-up to earlier research on the ADAPT, a factor analytic study was conducted. Method: One thousand and sixty six ADAPTs from clients with (suspected) intellectual disabilities or borderline intellectual functioning and 129 ADAPTs from participants from the general population were collected along with other characteristics (e.g., IQ, psychiatric classifications, living situation). Results: An exploratory factor analysis (EFA) was performed and resulted in good fit indices. Subsequent confirmatory factor analysis (CFA) and multigroup CFA showed acceptable to good fit indices. This resulted in an instrument with eight factors and 62 items. Conclusion: Factor analytic results suggest that the ADAPT is a valid instrument that measures adaptive skills in individuals with intellectual disabilities or borderline intellectual functioning. 10 p.

Bayesian estimation of multilevel structural equation models (MLSEMs) offers advantages in terms of sample size requirements and computational feasibility, but does require careful specification of the prior distribution especially for the random effects variance parameters. The traditional “non-informative” conjugate choice of an inverse-Gamma prior with small hyperparameters has been shown time and again to be problematic. In this paper, we investigate alternative, more robust prior distributions for the doubly latent categorical multilevel model. In contrast to multilevel models without latent variables, MLSEMs have multiple random effects variance parameters both for the multilevel structure and for the latent variable structure. It is therefore even more important to construct reasonable priors for these parameters. We find that, although the robust priors outperform the traditional inverse-Gamma prior, their hyperparameters do require careful consideration.

School closures due to the COVID-19 pandemic raised concerns about increases in educational inequality. We examined the magnitude of the impact of the first school closure for vulnerable student groups in particular. This study was conducted among 886 Grade 3 - 5 students in the Netherlands in schools serving a high percentage of students from disadvantaged backgrounds. Piecewise growth analyses indicated that the school closures caused discontinuity in students’ achievement growth on national standardized tests and led to an average learning loss of 2.47 months in mathematics and 2.35 in reading comprehension, exceeding the duration of the school closure. Findings suggest that school closures contribute to educational inequality and indicate which students may particularly need additional support to overcome the adverse consequences of the lockdowns.

Increasing research uses intensive longitudinal designs to examine antecedents and consequences associated with dynamic affective processes. These studies often rely on the Positive and Negative Affect Schedule (PANAS) to measure affect. Studies assessing the structure of the PANAS are largely cross-sectional in nature and cannot always disentangle within-person variability from between-person differences in affect. A paucity of studies examines structural similarities and differences in affect at the between and within-person levels, and few have done so with short-form versions of the PANAS. This study investigates the multilevel factor structure of the 10-item PANAS–short-form in a sample of young adults (n = 272) measured daily consecutively over 1 month. Additionally, dynamic relations between positive and negative affect, depressive symptoms, stress, and physical symptoms were examined. Results support a three factors within and two factors between multilevel structural model. Distinct dynamic relations were observed among positive affect, negative affect, stress, and physical symptoms at the within level. Positive and negative affect were correlated with depressive symptoms, stress, and physical symptoms at the between level. Findings indicate the need to disentangle structural components of positive and negative affect when conducting intensive longitudinal studies to examine correlates linked to dynamic affective processes.

Statistical network models based on Pairwise Markov Random Fields (PMRFs) are popular tools for analyzing multivariate psychological data, in large part due to their perceived role in generating insights into causal relationships: a practice known as causal discovery in the causal modelling literature. However, since network models are not presented as causal discovery tools, the role they play in generating causal insights is poorly understood among empirical researchers. In this paper we provide a treatment of how PMRFs such as the Gaussian Graphical Model (GGM) work as causal discovery tools, using Directed Acyclic Graphs (DAGs) and Structural Equation Models (SEMs) as causal models. We describe the key assumptions needed for causal discovery and show the equivalence class of causal models that networks identify from data. We clarify four common misconceptions found in the empirical literature relating to: networks as causal skeletons; chains of relationships; collider bias; and cyclic causal models.

Retrospective Assessment (RA) scores are often found to be higher than the mean of Ecological Momentary Assessment (EMA) scores about a concurrent period. This difference is generally interpreted as bias towards salient experiences in RA. During RA, participants are often asked to summarize their experiences in unspecific terms, which may indeed facilitate bias. At least in this unspecific form, the summary that participants apply to their remembered experiences can take many different forms. In this study, we reanalyzed an existing dataset (N = 92) using a repeated N = 1 approach. We reported on interindividual differences between EMA data and RA score, and assessed for each participant whether it was likely that their RA score was an approximation of the mean of their experiences as captured by their EMA data. We found considerable interpersonal differences in the difference between EMA scores and RA scores, as well as some extreme cases. Furthermore, for a considerable part of the sample (n = 46 for positive affect, n = 60 for negative affect), we did not reject the null hypothesis that their RA score represented the mean of their experiences as captured by their EMA data. We conclude that in its current unspecific form, RA may facilitate bias, although not for everyone. Future studies may determine whether more specific forms of RA reduce bias, while acknowledging interindividual differences.

Inferring reciprocal effects or causality between variables is a central aim of behavioral and psychological research. To address reciprocal effects, a variety of longitudinal models that include cross-lagged relations have been proposed in different contexts and disciplines. However, the relations between these cross-lagged models have not been systematically discussed in the literature. This lack of insight makes it difficult for researchers to select an appropriate model when analyzing longitudinal data, and some researchers do not even think about alternative cross-lagged models. The present research provides a unified framework that clarifies the conceptual and mathematical similarities and differences between these models. The unified framework shows that existing longitudinal models can be effectively classified based on whether the model posits unique factors and/or dynamic residuals and what types of common factors are used to model changes. The latter is essential to understand how cross-lagged parameters are interpreted. We also present an example using empirical data to demonstrate that there is great risk of drawing different conclusions depending on the cross-lagged models used. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Network analysis of ESM data has become popular in clinical psychology. In this approach, discrete-time (DT) vector auto-regressive (VAR) models define the network structure with centrality measures used to identify intervention targets. However, VAR models suffer from time-interval dependency. Continuous-time (CT) models have been suggested as an alternative but require a conceptual shift, implying that DT-VAR parameters reflect total rather than direct effects. In this paper, we propose and illustrate a CT network approach using CT-VAR models. We define a new network representation and develop centrality measures which inform intervention targeting. This methodology is illustrated with an ESM dataset.

Negation is a complex grammatical phenomenon that has received considerable attention in the biomedical natural language processing domain. While neural network-based methods are the state-of-the-art in negation scope resolution, they often use the unrealistic assumption that negation cue information is completely accurate. Even if this assumption holds, there remains a dependency on engineered features from state-of-the-art machine learning methods. To tackle this issue, in this study, we adopted a two-step negation resolving approach to assess whether a neural network-based model, here a bidirectional long short-term memory, can be a an alternative for cue detection. Furthermore, we investigate how inaccurate cue predictions would affect the scope resolution performance. We ran various experiments on the open access Bio-Scope corpus. Experimental results suggest that word embeddings alone can detect cues reasonably well, but there still exist better alternatives for this task. As expected, scope resolution performance suffers from imperfect cue information, but remains acceptable on the Abstracts subcorpus. We also found that the scope resolution performance is most robust against inaccurate information for models with a recurrent layer only, compared to extensions with a conditional random field layer and extensions with a post-processing algorithm. We advocate for more research into the application of automated deep learning on the effect of imperfect information on scope resolution.

Contains fulltext : 249813.pdf (Publisher’s version ) (Open Access) In recent years, network approaches to psychopathology have sparked much debate and have had a significant impact on how mental disorders are perceived in the field of clinical psychology. However, there are many important challenges in moving from theory to empirical research and clinical practice and vice versa. Therefore, in this article, we bring together different points of view on psychological networks by methodologists and clinicians to give a critical overview on these challenges, and to present an agenda for addressing these challenges. In contrast to previous reviews, we especially focus on methodological issues related to temporal networks. This includes topics such as selecting and assessing the quality of the nodes in the network, distinguishing between- and within-person effects in networks, relating items that are measured at different time scales, and dealing with changes in network structures. These issues are not only important for researchers using network models on empirical data, but also for clinicians, who are increasingly likely to encounter (person-specific) networks in the consulting room. 15 p.

This article describes some potential uses of Bayesian estimation for time-series and panel data models by incorporating information from prior probabilities (i.e., priors) in addition to observed data. Drawing on econometrics and other literatures we illustrate the use of informative "shrinkage" or "small variance" priors (including so-called "Minnesota priors") while extending prior work on the general cross-lagged panel model (GCLM). Using a panel dataset of national income and subjective well-being (SWB) we describe three key benefits of these priors. First, they shrink parameter estimates toward zero or toward each other for time-varying parameters, which lends additional support for an income → SWB effect that is not supported with maximum likelihood (ML). This is useful because, second, these priors increase model parsimony and the stability of estimates (keeping them within more reasonable bounds) and thus improve out-of-sample predictions and interpretability, which means estimated effect should also be more trustworthy than under ML. Third, these priors allow estimating otherwise under-identified models under ML, allowing higher-order lagged effects and time-varying parameters that are otherwise impossible to estimate using observed data alone. In conclusion we note some of the responsibilities that come with the use of priors which, departing from typical commentaries on their scientific applications, we describe as involving reflection on how best to apply modeling tools to address matters of worldly concern. ispartof: FRONTIERS IN PSYCHOLOGY vol:12 ispartof: location:Switzerland status: published

Background Comorbidity between depressive and anxiety disorders is common. A hypothesis of the network perspective on psychopathology is that comorbidity arises due to the interplay of symptoms shared by both disorders, with overlapping symptoms acting as so-called bridges, funneling symptom activation between symptom clusters of each disorder. This study investigated this hypothesis by testing whether (i) two overlapping mental states “worrying” and “feeling irritated” functioned as bridges in dynamic mental state networks of individuals with both depression and anxiety as compared to individuals with either disorder alone, and (ii) overlapping or non-overlapping mental states functioned as stronger bridges. Methods Data come from the Netherlands Study of Depression and Anxiety (NESDA). A total of 143 participants met criteria for comorbid depression and anxiety (65%), 40 participants for depression-only (18.2%), and 37 for anxiety-only (16.8%) during any NESDA wave. Participants completed momentary assessments of symptoms (i.e., mental states) of depression and anxiety, five times a day, for 2 weeks (14,185 assessments). First, dynamics between mental states were modeled with a multilevel vector autoregressive model, using Bayesian estimation. Summed average lagged indirect effects through the hypothesized bridge mental states were compared between groups. Second, we evaluated the role of all mental states as potential bridge mental states. Results While the summed indirect effect for the bridge mental state “worrying” was larger in the comorbid group compared to the single disorder groups, differences between groups were not statistically significant. The difference between groups became more pronounced when only examining individuals with recent diagnoses ( Conclusions This study empirically examined a prominent network-approach hypothesis for the first time using longitudinal data. No support was found for overlapping mental states “worrying” and “feeling irritable” functioning as bridge mental states in individuals vulnerable for comorbid depression and anxiety. Potentially, bridge mental state activity can only be observed during acute symptomatology. If so, these may present as interesting targets in treatment, but not prevention. This requires further investigation.

Ecological Momentary Assessment (EMA) in which participants report on their moment-to-moment experiences in their natural environment, is a hot topic. An emerging field in clinical psychology based on either EMA, or what we term Ecological Retrospective Assessment (ERA) as it requires retrospectivity, is the field of personalized feedback. In this field, EMA/ERA-data-driven summaries are presented to participants with the goal of promoting their insight in their experiences. Underlying this procedure are some fundamental assumptions about (i) the relation between true moment-to-moment experiences and retrospective evaluations of those experiences, (ii) the translation of these experiences and evaluations to different types of data, (iii) the comparison of these different types of data, and (iv) the impact of a summary of moment-to-moment experiences on retrospective evaluations of those experiences. We argue that these assumptions deserve further exploration, in order to create a strong evidence-based foundation for the personalized feedback procedure.

The random intercept cross-lagged panel model (RI-CLPM) is rapidly gaining popularity in psychology and related fields as a structural equation modeling (SEM) approach to longitudinal data. It decomposes observed scores into within-unit dynamics and stable, between-unit differences. This paper discusses three extensions of the RI-CLPM that researchers may be interested in, but are unsure of how to accomplish: (a) including stable, person-level characteristics as predictors and/or outcomes; (b) specifying a multiple-group version; and (c) including multiple indicators. For each extension, we discuss which models need to be run in order to investigate underlying assumptions, and we demonstrate the various modeling options using a motivating example. We provide fully annotated code for lavaan (R-package) and Mplus on an accompanying website.

In recent years, a growing chorus of researchers have argued that psychological theory is in a state of crisis: theories are rarely developed in a way that indicates an accumulation of knowledge and they are often absent from our research entirely. More than 40 years ago, Paul Meehl raised these very concerns. Yet, in the ensuing decades, little has improved. We aim to chart a better path forward for psychological theory by revisiting Meehl's criticisms, his proposed solution, and the reasons his solution failed to meaningful change the status of psychological theory. We argue that Meehl identified serious shortcomings in our evaluation of psychological theories and that his proposed solution would substantially strengthen theory testing. However, we also argue that he failed to provide researchers a set of tools for theory construction. To advance psychological theory, we must equip researchers with tools to better generate, evaluate, and develop their theories. We argue that formal theories provide this much needed set of tools, equipping researchers with tools for thinking, evaluating explanation, informing theory development, strengthening measurement, and moving toward collaborative construction of psychological theories that allow us to explain, predict, and control psychological phenomena.

There has been a strong increase in the number of studies based on intensive longitudinal data, such as those obtained with experience sampling and daily diaries. These data contain a wealth of information regarding the dynamics of processes as they unfold within individuals over time. In this article, we discuss how combining intensive longitudinal data with either time-series analysis, which consists of modeling the temporal dependencies in the data for a single individual, or dynamic multilevel modeling, which consists of using a time-series model at Level 1 to describe the within-person process while allowing for individual differences in the parameters of these processes at Level 2, has led to new insights in clinical psychology. In addition, we discuss several methodological and statistical challenges that researchers face when they are interested in studying the dynamics of psychological processes using intensive longitudinal data.

Complexity science methods offer new opportunities for prognosis and treatment in healthcare and clinical psychology because of the increasing need for integration of the detailed knowledge of physiological and psychological subsystems and the increasing prevalence of multiple disease conditions in our aging societies. This chapter explains how the frequently occurring acute transitions and related tipping points in physical and mental processes in these populations can be monitored with time series and dynamical indicators of resilience. The authors introduce slowing down of recovery, increase in variance and autocorrelation, and increasing cross-correlation between subsystem time series as valid predictors of the proximity of tipping points in diseases such as depression, heart failure and syncope. Using wearable devices, together with these complex systems analyses, yields new methods of forecasting and may improve resilience of individual persons and their mental or physical (organ) subsystems