Using clinical expertise and empirical data in constructing networks of trauma symptoms in refugee youth

Background: In recent years, many adolescents have fled their home countries due to war and human rights violations, consequently experiencing various traumatic events and putting them at risk of developing mental health problems. The symptomatology of refugee youth was shown to be multifaceted and often falling outside of traditional diagnoses. Objective: The present study aimed to investigate the symptomatology of this patient group by assessing the network structure of a wide range of symptoms. Further, we assessed clinicians' perceptions of symptoms relations in order to evaluate the clinical validity of the empirical network. Methods: Empirical data on Post-Traumatic Stress Disorder (PTSD), depression and other trauma symptoms from N = 366 refugee youth were collected during the routine diagnostic process of an outpatient centre for refugee youth in Germany. Additionally, four clinicians of this outpatient centre were asked how they perceive symptom relations in their patients using a newly developed tool. Separate networks were constructed based on 1) empirical symptom data and 2) clinicians' perceived symptom relations (PSR). Results: Both the network based on empirical data and the network based on clinicians' PSR showed that symptoms of PTSD and depression related most strongly within each respective cluster (connected mainly via sleeping problems), externalizing symptoms were somewhat related to PTSD symptoms and intrusions were central. Some differences were found within the clinicians' PSR as well as between the PSR and the empirical network. Still, the general PSR-network structure showed a moderate to good fit to the empirical data. Conclusion: Our results suggest that sleeping problems and intrusions play a central role in the symptomatology of refugee children, which has tentative implications for diagnostics and treatment. Further, externalizing symptoms might be an indicator for PTSD-symptoms. Finally, using clinicians' PSR for network construction offered a promising possibility to gain information on symptom networks and their clinical validity.

Interpreting the Ising Model: The Input Matters

The Ising model is a model for pairwise interactions between binary variables that has become popular in the psychological sciences. It has been first introduced as a theoretical model for the alignment between positive (1) and negative (-1) atom spins. In many psychological applications, however, the Ising model is defined on the domain {0, 1} instead of the classical domain {-1,1}. While it is possible to transform the parameters of the Ising model in one domain to obtain a statistically equivalent model in the other domain, the parameters in the two versions of the Ising model lend themselves to different interpretations and imply different dynamics, when studying the Ising model as a dynamical system. In this tutorial paper, we provide an accessible discussion of the interpretation of threshold and interaction parameters in the two domains and show how the dynamics of the Ising model depends on the choice of domain. Finally, we provide a transformation that allows one to transform the parameters in an Ising model in one domain into a statistically equivalent Ising model in the other domain.

Investigating the Utility of Fixed-margin Sampling in Network Psychometrics

Steinley, Hoffman, Brusco, and Sher (2017) proposed a new method for evaluating the performance of psychological network models: fixed-margin sampling. The authors investigated LASSO regularized Ising models (eLasso) by generating random datasets with the same margins as the original binary dataset, and concluded that many estimated eLasso parameters are not distinguishable from those that would be expected if the data were generated by chance. We argue that fixed-margin sampling cannot be used for this purpose, as it generates data under a particular null-hypothesis: a unidimensional factor model with interchangeable indicators (i.e., the Rasch model). We show this by discussing relevant psychometric literature and by performing simulation studies. Results indicate that while eLasso correctly estimated network models and estimated almost no edges due to chance, fixed-margin sampling performed poorly in classifying true effects as "interesting" (Steinley et al. 2017, p. 1004). Further simulation studies indicate that fixed-margin sampling offers a powerful method for highlighting local misfit from the Rasch model, but performs only moderately in identifying global departures from the Rasch model. We conclude that fixed-margin sampling is not up to the task of assessing if results from estimated Ising models or other multivariate psychometric models are due to chance.

On the Importance of Estimating Parameter Uncertainty in Network Psychometrics: A Response to Forbes et al. (2019)

In their recent paper, Forbes et al. (2019; FWMK) evaluate the replicability of network models in two studies. They identify considerable replicability issues, concluding that "current 'state-of-the-art' methods in the psychopathology network literature […] are not well-suited to analyzing the structure of the relationships between individual symptoms". Such strong claims require strong evidence, which the authors do not provide. FWMK identify low replicability by analyzing point estimates of networks; contrast low replicability with results of two statistical tests that indicate higher replicability, and conclude that these tests are problematic. We make four points. First, statistical tests are superior to the visual inspection of point estimates, because tests take into account sampling variability. Second, FWMK misinterpret the statistical tests in several important ways. Third, FWMK did not follow established recommendations when estimating networks in their first study, underestimating replicability. Fourth, FWMK draw conclusions about methodology, which does not follow from investigations of data, and requires investigations of methodology. Overall, we show that the "poor replicability "observed by FWMK occurs due to sampling variability and use of suboptimal methods. We conclude by discussing important recent simulation work that guides researchers to use models appropriate for their data, such as nonregularized estimation routines.

Psychological networks in clinical populations: investigating the consequences of Berkson's bias

BACKGROUND

In clinical research, populations are often selected on the sum-score of diagnostic criteria such as symptoms. Estimating statistical models where a subset of the data is selected based on a function of the analyzed variables introduces Berkson's bias, which presents a potential threat to the validity of findings in the clinical literature. The aim of the present paper is to investigate the effect of Berkson's bias on the performance of the two most commonly used psychological network models: the Gaussian Graphical Model (GGM) for continuous and ordinal data, and the Ising Model for binary data.

METHODS

In two simulation studies, we test how well the two models recover a true network structure when estimation is based on a subset of the data typically seen in clinical studies. The network is based on a dataset of 2807 patients diagnosed with major depression, and nodes in the network are items from the Hamilton Rating Scale for Depression (HRSD). The simulation studies test different scenarios by varying (1) sample size and (2) the cut-off value of the sum-score which governs the selection of participants.

RESULTS

The results of both studies indicate that higher cut-off values are associated with worse recovery of the network structure. As expected from the Berkson's bias literature, selection reduced recovery rates by inducing negative connections between the items.

CONCLUSION

Our findings provide evidence that Berkson's bias is a considerable and underappreciated problem in the clinical network literature. Furthermore, we discuss potential solutions to circumvent Berkson's bias and their pitfalls.

Personalized networks of eating disorder symptoms predicting eating disorder outcomes and remission

Enhanced cognitive-behavioral therapy (CBT-E) is one of the primary evidence-based treatments for adults with eating disorders (EDs). However, up to 50% of individuals do not respond to CBT-E, likely because of the high heterogeneity present even within similar diagnoses. This high heterogeneity, especially in regard to presenting pathology, makes it difficult to develop a treatment based "on averages" and for clinicians to accurately pinpoint which symptoms should be targeted in treatment. As such, new models based at both the group, and individual level, are needed to more accurately refine targets for personalized evidence-based treatments that can lead to full remission. The current study (Expected N = 120 anorexia nervosa, atypical anorexia nervosa, and bulimia nervosa) will build both group and individual longitudinal models of ED behaviors, cognitions, affect, and physiology. We will collect data for 30 days utilizing a mobile application to assess behaviors, cognition, and affect and a sensor wristband that assesses physiology (heart rate, acceleration). We will also collect outcome data at 1- and 6-month follow-ups to assess ED outcomes and remission status. These data will allow for identification of "on average" and "individual" targets that maintain ED pathology and test if these targets predict outcomes, including ED remission.

How to Compare Psychometric Factor and Network Models

In memory of Dr. Dennis John McFarland, who passed away recently, our objective is to continue his efforts to compare psychometric networks and latent variable models statistically. We do so by providing a commentary on his latest work, which he encouraged us to write, shortly before his death. We first discuss the statistical procedure McFarland used, which involved structural equation modeling (SEM) in standard SEM software. Next, we evaluate the penta-factor model of intelligence. We conclude that (1) standard SEM software is not suitable for the comparison of psychometric networks with latent variable models, and (2) the penta-factor model of intelligence is only of limited value, as it is nonidentified. We conclude with a reanalysis of the Wechlser Adult Intelligence Scale data McFarland discussed and illustrate how network and latent variable models can be compared using the recently developed R package Psychonetrics. Of substantive theoretical interest, the results support a network interpretation of general intelligence. A novel empirical finding is that networks of intelligence replicate over standardization samples.

Time to get personal? The impact of researchers choices on the selection of treatment targets using the experience sampling methodology

OBJECTIVE

One of the promises of the experience sampling methodology (ESM) is that a statistical analysis of an individual's emotions, cognitions and behaviors in everyday-life could be used to identify relevant treatment targets. A requisite for clinical implementation is that outcomes of such person-specific time-series analyses are not wholly contingent on the researcher performing them.

METHODS

To evaluate this, we crowdsourced the analysis of one individual patient's ESM data to 12 prominent research teams, asking them what symptom(s) they would advise the treating clinician to target in subsequent treatment.

RESULTS

Variation was evident at different stages of the analysis, from preprocessing steps (e.g., variable selection, clustering, handling of missing data) to the type of statistics and rationale for selecting targets. Most teams did include a type of vector autoregressive model, examining relations between symptoms over time. Although most teams were confident their selected targets would provide useful information to the clinician, not one recommendation was similar: both the number (0-16) and nature of selected targets varied widely.

CONCLUSION

This study makes transparent that the selection of treatment targets based on personalized models using ESM data is currently highly conditional on subjective analytical choices and highlights key conceptual and methodological issues that need to be addressed in moving towards clinical implementation.

Bridging the gap between complexity science and clinical practice by formalizing idiographic theories: a computational model of functional analysis

BACKGROUND

The past decades of research have seen an increase in statistical tools to explore the complex dynamics of mental health from patient data, yet the application of these tools in clinical practice remains uncommon. This is surprising, given that clinical reasoning, e.g., case conceptualizations, largely coincides with the dynamical system approach. We argue that the gap between statistical tools and clinical practice can partly be explained by the fact that current estimation techniques disregard theoretical and practical considerations relevant to psychotherapy. To address this issue, we propose that case conceptualizations should be formalized. We illustrate this approach by introducing a computational model of functional analysis, a framework commonly used by practitioners to formulate case conceptualizations and design patient-tailored treatment.

METHODS

We outline the general approach of formalizing idiographic theories, drawing on the example of a functional analysis for a patient suffering from panic disorder. We specified the system using a series of differential equations and simulated different scenarios; first, we simulated data without intervening in the system to examine the effects of avoidant coping on the development of panic symptomatic. Second, we formalized two interventions commonly used in cognitive behavioral therapy (CBT; exposure and cognitive reappraisal) and subsequently simulated their effects on the system.

RESULTS

The first simulation showed that the specified system could recover several aspects of the phenomenon (panic disorder), however, also showed some incongruency with the nature of panic attacks (e.g., rapid decreases were not observed). The second simulation study illustrated differential effects of CBT interventions for this patient. All tested interventions could decrease panic levels in the system.

CONCLUSIONS

Formalizing idiographic theories is promising in bridging the gap between complexity science and clinical practice and can help foster more rigorous scientific practices in psychotherapy, through enhancing theory development. More precise case conceptualizations could potentially improve intervention planning and treatment outcomes. We discuss applications in psychotherapy and future directions, amongst others barriers for systematic theory evaluation and extending the framework to incorporate interactions between individual systems, relevant for modeling social learning processes. With this report, we hope to stimulate future efforts in formalizing clinical frameworks.

Bridging the gap between complexity science and clinical practice by formalizing idiographic theories: a computational model of functional analysis

BACKGROUND

The past decades of research have seen an increase in statistical tools to explore the complex dynamics of mental health from patient data, yet the application of these tools in clinical practice remains uncommon. This is surprising, given that clinical reasoning, e.g., case conceptualizations, largely coincides with the dynamical system approach. We argue that the gap between statistical tools and clinical practice can partly be explained by the fact that current estimation techniques disregard theoretical and practical considerations relevant to psychotherapy. To address this issue, we propose that case conceptualizations should be formalized. We illustrate this approach by introducing a computational model of functional analysis, a framework commonly used by practitioners to formulate case conceptualizations and design patient-tailored treatment.

METHODS

We outline the general approach of formalizing idiographic theories, drawing on the example of a functional analysis for a patient suffering from panic disorder. We specified the system using a series of differential equations and simulated different scenarios; first, we simulated data without intervening in the system to examine the effects of avoidant coping on the development of panic symptomatic. Second, we formalized two interventions commonly used in cognitive behavioral therapy (CBT; exposure and cognitive reappraisal) and subsequently simulated their effects on the system.

RESULTS

The first simulation showed that the specified system could recover several aspects of the phenomenon (panic disorder), however, also showed some incongruency with the nature of panic attacks (e.g., rapid decreases were not observed). The second simulation study illustrated differential effects of CBT interventions for this patient. All tested interventions could decrease panic levels in the system.

CONCLUSIONS

Formalizing idiographic theories is promising in bridging the gap between complexity science and clinical practice and can help foster more rigorous scientific practices in psychotherapy, through enhancing theory development. More precise case conceptualizations could potentially improve intervention planning and treatment outcomes. We discuss applications in psychotherapy and future directions, amongst others barriers for systematic theory evaluation and extending the framework to incorporate interactions between individual systems, relevant for modeling social learning processes. With this report, we hope to stimulate future efforts in formalizing clinical frameworks.

Toward incorporating genetic risk scores into symptom networks of psychosis

BACKGROUND

Psychosis spectrum disorder is a heterogeneous, multifactorial clinical phenotype, known to have a high heritability, only a minor portion of which can be explained by molecular measures of genetic variation. This study proposes that the identification of genetic variation underlying psychotic disorder may have suffered due to issues in the psychometric conceptualization of the phenotype. Here we aim to open a new line of research into the genetics of mental disorders by explicitly incorporating genes into symptom networks. Specifically, we investigate whether links between a polygenic risk score (PRS) for schizophrenia and measures of psychosis proneness can be identified in a network model.

METHODS

We analyzed data from n = 2180 subjects (controls, patients diagnosed with a non-affective psychotic disorder, and the first-degree relatives of the patients). A network structure was computed to examine associations between the 42 symptoms of the Community Assessment of Psychic Experiences (CAPE) and the PRS for schizophrenia.

RESULTS

The resulting network shows that the PRS is directly connected to the spectrum of positive and depressive symptoms, with the items conspiracy and no future being more often located on predictive pathways from PRS to other symptoms.

CONCLUSIONS

To our knowledge, the current exploratory study provides a first application of the network framework to the field of behavior genetics research. This allows for a novel outlook on the investigation of the relations between genome-wide association study-based PRSs and symptoms of mental disorders, by focusing on the dependencies among variables.

Psychometric network models from time-series and panel data

Researchers in the field of network psychometrics often focus on the estimation of Gaussian graphical models (GGMs)-an undirected network model of partial correlations-between observed variables of cross-sectional data or single-subject time-series data. This assumes that all variables are measured without measurement error, which may be implausible. In addition, cross-sectional data cannot distinguish between within-subject and between-subject effects. This paper provides a general framework that extends GGM modeling with latent variables, including relationships over time. These relationships can be estimated from time-series data or panel data featuring at least three waves of measurement. The model takes the form of a graphical vector-autoregression model between latent variables and is termed the ts-lvgvar when estimated from time-series data and the panel-lvgvar when estimated from panel data. These methods have been implemented in the software package psychonetrics, which is exemplified in two empirical examples, one using time-series data and one using panel data, and evaluated in two large-scale simulation studies. The paper concludes with a discussion on ergodicity and generalizability. Although within-subject effects may in principle be separated from between-subject effects, the interpretation of these results rests on the intensity and the time interval of measurement and on the plausibility of the assumption of stationarity.

What do centrality measures measure in psychological networks?

Centrality indices are a popular tool to analyze structural aspects of psychological networks. As centrality indices were originally developed in the context of social networks, it is unclear to what extent these indices are suitable in a psychological network context. In this article we critically examine several issues with the use of the most popular centrality indices in psychological networks: degree, betweenness, and closeness centrality. We show that problems with centrality indices discussed in the social network literature also apply to the psychological networks. Assumptions underlying centrality indices, such as presence of a flow and shortest paths, may not correspond with a general theory of how psychological variables relate to one another. Furthermore, the assumptions of node distinctiveness and node exchangeability may not hold in psychological networks. We conclude that, for psychological networks, betweenness and closeness centrality seem especially unsuitable as measures of node importance. We therefore suggest three ways forward: (a) using centrality measures that are tailored to the psychological network context, (b) reconsidering existing measures of importance used in statistical models underlying psychological networks, and (c) discarding the concept of node centrality entirely. Foremost, we argue that one has to make explicit what one means when one states that a node is central, and what assumptions the centrality measure of choice entails, to make sure that there is a match between the process under study and the centrality measure that is used. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Exploring multicultural classroom dynamics: A network analysis

Students' relationships with peers and teachers strongly influence their motivation to engage in learning activities. Ethnic minority students, however, are often victimized in schools, and their educational achievement lags behind that of their majority group counterparts. The aim of the present study was to explore teachers' multicultural approach within their classrooms as a possible factor of influence over students' peer relationships and motivation. We utilized the novel methodology of estimating psychological networks in order to map out the interactions between these constructs within multicultural classrooms. Results indicate that a multicultural approach is directly connected to student motivation for both ethnic majority and minority students. Social integration within peer groups, however, seems to be a possible mediator of this relationship for the ethnic minority students. Due to the hypothesis generating nature of the psychological network approach, a more thorough investigation of this generated mediation hypothesis is called for.

Reproducibility and Replicability in a Fast-Paced Methodological World

Methodological developments and software implementations are progressing at an increasingly fast pace. The introduction and widespread acceptance of preprint archived reports and open-source software have made state-of-the-art statistical methods readily accessible to researchers. At the same time, researchers are increasingly concerned that their results should be reproducible (i.e., the same analysis should yield the same numeric results at a later time), which is a basic requirement for assessing the results’ replicability (i.e., whether results at a later time support the same conclusions). Although this age of fast-paced methodology greatly facilitates reproducibility and replicability, it also undermines them in ways not often realized by researchers. This article draws researchers’ attention to these threats and proposes guidelines to help minimize their impact. Reproducibility may be influenced by software development and change over time, a problem that is greatly compounded by the rising dependency between software packages. Replicability is affected by rapidly changing standards, researcher degrees of freedom, and possible bugs or errors in code, whether introduced by software developers or empirical researchers implementing an analysis. This article concludes with a list of recommendations to improve the reproducibility and replicability of results.

A tutorial on regularized partial correlation networks

Recent years have seen an emergence of network modeling applied to moods, attitudes, and problems in the realm of psychology. In this framework, psychological variables are understood to directly affect each other rather than being caused by an unobserved latent entity. In this tutorial, we introduce the reader to estimating the most popular network model for psychological data: the partial correlation network. We describe how regularization techniques can be used to efficiently estimate a parsimonious and interpretable network structure in psychological data. We show how to perform these analyses in R and demonstrate the method in an empirical example on posttraumatic stress disorder data. In addition, we discuss the effect of the hyperparameter that needs to be manually set by the researcher, how to handle non-normal data, how to determine the required sample size for a network analysis, and provide a checklist with potential solutions for problems that can arise when estimating regularized partial correlation networks. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

A tutorial on regularized partial correlation networks

Recent years have seen an emergence of network modeling applied to moods, attitudes, and problems in the realm of psychology. In this framework, psychological variables are understood to directly affect each other rather than being caused by an unobserved latent entity. In this tutorial, we introduce the reader to estimating the most popular network model for psychological data: the partial correlation network. We describe how regularization techniques can be used to efficiently estimate a parsimonious and interpretable network structure in psychological data. We show how to perform these analyses in R and demonstrate the method in an empirical example on posttraumatic stress disorder data. In addition, we discuss the effect of the hyperparameter that needs to be manually set by the researcher, how to handle non-normal data, how to determine the required sample size for a network analysis, and provide a checklist with potential solutions for problems that can arise when estimating regularized partial correlation networks. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

The Network Structure of Schizotypal Personality Traits

Elucidating schizotypal traits is important if we are to understand the various manifestations of psychosis spectrum liability and to reliably identify individuals at high risk for psychosis. The present study examined the network structures of (1) 9 schizotypal personality domains and (2) 74 individual schizotypal items, and (3) explored whether networks differed across gender and culture (North America vs China). The study was conducted in a sample of 27001 participants from 12 countries and 21 sites (M age = 22.12; SD = 6.28; 37.5% males). The Schizotypal Personality Questionnaire (SPQ) was used to assess 74 self-report items aggregated in 9 domains. We used network models to estimate conditional dependence relations among variables. In the domain-level network, schizotypal traits were strongly interconnected. Predictability (explained variance of each node) ranged from 31% (odd/magical beliefs) to 55% (constricted affect), with a mean of 43.7%. In the item-level network, variables showed relations both within and across domains, although within-domain associations were generally stronger. The average predictability of SPQ items was 27.8%. The network structures of men and women were similar (r = .74), node centrality was similar across networks (r = .90), as was connectivity (195.59 and 199.70, respectively). North American and Chinese participants networks showed lower similarity in terms of structure (r = 0.44), node centrality (r = 0.56), and connectivity (180.35 and 153.97, respectively). In sum, the present article points to the value of conceptualizing schizotypal personality as a complex system of interacting cognitive, emotional, and affective characteristics.

Dynamic networks of PTSD symptoms during conflict

BACKGROUND

Conceptualizing posttraumatic stress disorder (PTSD) symptoms as a dynamic system of causal elements could provide valuable insights into the way that PTSD develops and is maintained in traumatized individuals. We present the first study to apply a multilevel network model to produce an exploratory empirical conceptualization of dynamic networks of PTSD symptoms, using data collected during a period of conflict.

METHODS

Intensive longitudinal assessment data were collected during the Israel-Gaza War in July-August 2014. The final sample (n = 96) comprised a general population sample of Israeli adult civilians exposed to rocket fire. Participants completed twice-daily reports of PTSD symptoms via smartphone for 30 days. We used a multilevel vector auto-regression model to produce contemporaneous and temporal networks, and a partial correlation network model to obtain a between-subjects network.

RESULTS

Multilevel network analysis found strong positive contemporaneous associations between hypervigilance and startle response, avoidance of thoughts and avoidance of reminders, and between flashbacks and emotional reactivity. The temporal network indicated the central role of startle response as a predictor of future PTSD symptomatology, together with restricted affect, blame, negative emotions, and avoidance of thoughts. There were some notable differences between the temporal and contemporaneous networks, including the presence of a number of negative associations, particularly from blame. The between-person network indicated flashbacks and emotional reactivity to be the most central symptoms.

CONCLUSIONS

This study suggests various symptoms that could potentially be driving the development of PTSD. We discuss clinical implications such as identifying particular symptoms as targets for interventions.

The Gaussian Graphical Model in Cross-Sectional and Time-Series Data

We discuss the Gaussian graphical model (GGM; an undirected network of partial correlation coefficients) and detail its utility as an exploratory data analysis tool. The GGM shows which variables predict one-another, allows for sparse modeling of covariance structures, and may highlight potential causal relationships between observed variables. We describe the utility in three kinds of psychological data sets: data sets in which consecutive cases are assumed independent (e.g., cross-sectional data), temporally ordered data sets (e.g., n = 1 time series), and a mixture of the 2 (e.g., n > 1 time series). In time-series analysis, the GGM can be used to model the residual structure of a vector-autoregression analysis (VAR), also termed graphical VAR. Two network models can then be obtained: a temporal network and a contemporaneous network. When analyzing data from multiple subjects, a GGM can also be formed on the covariance structure of stationary means-the between-subjects network. We discuss the interpretation of these models and propose estimation methods to obtain these networks, which we implement in the R packages graphicalVAR and mlVAR. The methods are showcased in two empirical examples, and simulation studies on these methods are included in the supplementary materials.