A structural after measurement approach to structural equation modeling

The core tendencies underlying prosocial behavior: Testing a person-situation framework

OBJECTIVE AND BACKGROUND

According to a recently proposed theoretical framework, different personality traits should explain pro-social behavior in different situations. We empirically tested the key proposition of this framework that each of four "core tendencies" (i.e., the shared variance of related traits) specifically predicts pro-social behavior in the presence of a different situational affordance.

METHODS

We used a large-scale dataset (N = 2479) including measures of various personality traits and six incentivized economic games assessing pro-social behavior in different social situations. Using bifactor modeling, we extracted four latent core tendencies and tested their predictive validity for pro-social behavior.

RESULTS

We found mixed support for the theoretically derived, preregistered hypotheses. The core tendency of beliefs about others' pro-sociality predicted pro-social behavior in both games involving dependence under uncertainty, as expected. Unconditional concern for others' welfare predicted pro-social behavior in only one of two games providing a possibility for exploitation. For conditional concern for others' welfare and self-regulation, in turn, evidence relating them to pro-social behavior in the presence of a possibility for reciprocity and temporal conflict was relatively weak.

CONCLUSION

Different features of social situations may activate different personality traits to influence pro-social behavior, but more research is needed to fully understand these person-situation interactions.

Somatic symptoms among young adults: an observational study examining the roles of trauma type and psychological distress

BACKGROUND

This study extends previous research examining the interplay between trauma and somatic symptoms by focusing on trauma type (i.e., whether the trauma was interpersonal in nature [e.g., assault, sexual violence, combat] or not) and the extent to which psychological distress accounts for these associations. Additionally, we novelly focus on clusters of somatic symptoms.

METHODS

A sample of predominantly Hispanic/Latinx young adults (n = 214) completed a series of brief validated questionnaires assessing demographics, stressful life events (Stressful Life Events Screening Questionnaire - revised), somatic symptoms (Patient Health Questionnaire - 14), and psychological distress (Patient Health Questionnaire - 4) as part of an online survey. Data were first analyzed using linear regression, followed by structural equation modeling to estimate indirect effects, with bootstrapping used to generate confidence intervals.

RESULTS

Results support a significant indirect effect of interpersonal trauma (IP) on somatic symptoms through psychological distress. While both IP and psychological distress contributed to cardiopulmonary and pain/fatigue clusters, gastrointestinal symptoms were accounted for by psychological distress. Exploratory analyses revealed unique associations by gender, with partial mediation of associations between IP and somatic symptoms by psychological distress observed more clearly in women.

CONCLUSIONS

The present study extends extant research demonstrating that greater exposure to interpersonal trauma exposure is significantly and strongly associated with increased somatic symptoms, psychological distress partially accounts for these associations. With replication, these findings inform theoretical frameworks of the psychological underpinnings of somatic symptom development and can be used to foster advancements in patient care.

Latently Mediating: A Bayesian Take on Causal Mediation Analysis with Structured Survey Data

In this paper, we propose a Bayesian causal mediation approach to the analysis of experimental data when both the outcome and the mediator are measured through structured questionnaires based on Likert-scaled inquiries. Our estimation strategy builds upon the error-in-variables literature and, specifically, it leverages Item Response Theory to explicitly model the observed surrogate mediator and outcome measures. We employ their elicited latent counterparts in a simple g-computation algorithm, where we exploit the fundamental identifying assumptions of causal mediation analysis to impute all the relevant counterfactuals and estimate the causal parameters of interest. We finally devise a sensitivity analysis procedure to test the robustness of the proposed methods to the restrictive requirement of mediator's conditional ignorability. We demonstrate the functioning of our proposed methodology through an empirical application using survey data from an online experiment on food purchasing intentions and the effect of different labeling regimes.

Depressive and Negative Symptoms in the Early and Established Stages of Schizophrenia: Integrating Structural Brain Alterations, Cognitive Performance, and Plasma Interleukin 6 Levels

Background

Depressive and negative symptoms are related to poor functional outcomes in schizophrenia. Cognitive deficits, reduced brain cortical thickness and volumes, and inflammation may contribute to depressive and negative symptoms, but pharmacological treatment and disease progression may confound the associations.

Methods

We evaluated whether higher plasma interleukin 6 (IL-6) levels would be associated with more severe negative or depressive symptoms in schizophrenia and explored illness stage utilizing early (BeneMin [Benefit of Minocycline on Negative Symptoms of Psychosis: Extent and Mechanism], n = 201, 72.8% male) and established (iRELATE [Immune Response & Social Cognition in Schizophrenia], n = 94, 67.3% male) schizophrenia cohorts. Using structural equation modeling in a subsample (iRELATE: n = 42, 69.0% male; BeneMin: n = 102, 76.5% male) with data on structural brain metrics (cortical thickness and volume), general cognitive performance, and plasma IL-6 levels, we assessed the interrelationships between these variables on depressive and negative symptom severity in early and established schizophrenia samples combined and in early schizophrenia only. All analyses were adjusted for sex, age, and chlorpromazine equivalent dose.

Results

Higher plasma IL-6 levels were related to more severe depressive symptoms in early schizophrenia (p < .05) and negative symptoms in established schizophrenia (p < .05). Structural equation modeling findings in early and established schizophrenia samples combined and early schizophrenia only showed that the interrelationship between higher plasma IL-6 levels, structural brain metrics, and general cognitive performance did not predict the severity of depressive and negative symptoms (p > .05). Higher plasma IL-6 levels and lower general cognitive performance were associated with reduced brain metrics (p < .05).

Conclusions

Our results indicate that higher plasma IL-6 levels may be differently associated with the severity of depressive and negative symptoms dependent on the illness stage. Future work identifying elevated levels of inflammation in larger samples may allow stratification and personalized intervention by subgroups who are at risk of poor outcomes.

Structural after measurement (SAM) approaches for accommodating latent quadratic and interaction effects

Established strategies commonly used to address latent quadratic and interaction effects within structural equation models, such as the unconstrained product indicator (UPI) approach or the latent moderated structural equations (LMS) approach, tend to perform effectively in models featuring only a limited number of nonlinear effects. However, as the complexity of the model increases with a higher number of nonlinear terms, the feasibility of joint or one-step methods such as UPI and LMS progressively diminishes. In response to this challenge, this paper advocates the adoption of structural after measurement (SAM) approaches to overcome this limitation. In a SAM approach, estimation proceeds in two stages. In a first stage, we estimate the parameters related to the measurement part of the model, while in a second stage, we estimate the parameters related to the structural part of the model. In this paper, we discuss three SAM approaches already documented in the literature and introduce a novel method based on the local SAM approach. To illustrate the utility of these SAM approaches, we conduct a modest simulation study, demonstrating that SAM approaches for latent quadratic and interaction effects offer a practical and viable alternative to the well-established one-step approaches.

Interactions between latent variables in count regression models

In psychology and the social sciences, researchers often model count outcome variables accounting for latent predictors and their interactions. Even though neglecting measurement error in such count regression models (e.g., Poisson or negative binomial regression) can have unfavorable consequences like attenuation bias, such analyses are often carried out in the generalized linear model (GLM) framework using fallible covariates such as sum scores. An alternative is count regression models based on structural equation modeling, which allow to specify latent covariates and thereby account for measurement error. However, the issue of how and when to include interactions between latent covariates or between latent and manifest covariates is rarely discussed for count regression models. In this paper, we present a latent variable count regression model (LV-CRM) allowing for latent covariates as well as interactions among both latent and manifest covariates. We conducted three simulation studies, investigating the estimation accuracy of the LV-CRM and comparing it to GLM-based count regression models. Interestingly, we found that even in scenarios with high reliabilities, the regression coefficients from a GLM-based model can be severely biased. In contrast, even for moderate sample sizes, the LV-CRM provided virtually unbiased regression coefficients. Additionally, statistical inferences yielded mixed results for the GLM-based models (i.e., low coverage rates, but acceptable empirical detection rates), but were generally acceptable using the LV-CRM. We provide an applied example from clinical psychology illustrating how the LV-CRM framework can be used to model count regressions with latent interactions.

Estimating power in complex nonlinear structural equation modeling including moderation effects: The powerNLSEM R-package

The model-implied simulation-based power estimation (MSPE) approach is a new general method for power estimation (Irmer et al., 2024). MSPE was developed especially for power estimation of non-linear structural equation models (SEM), but it also can be applied to linear SEM and manifest models using the R package powerNLSEM. After first providing some information about MSPE and the new adaptive algorithm that automatically selects sample sizes for the best prediction of power using simulation, a tutorial on how to conduct the MSPE for quadratic and interaction SEM (QISEM) using the powerNLSEM package is provided. Power estimation is demonstrated for four methods, latent moderated structural equations (LMS), the unconstrained product indicator (UPI), a simple factor score regression (FSR), and a scale regression (SR) approach to QISEM. In two simulation studies, we highlight the performance of the MSPE for all four methods applied to two QISEM with varying complexity and reliability. Further, we justify the settings of the newly developed adaptive search algorithm via performance evaluations using simulation. Overall, the MSPE using the adaptive approach performs well in terms of bias and Type I error rates.

Deriving Expected Values of Model Parameters when Using Sum Scores in Simulation Research

There is increasing interest in using factor scores in structural equation models and there have been numerous methodological papers on the topic. Nevertheless, sum scores, which are computed from adding up item responses, continue to be ubiquitous in practice. It is therefore important to compare simulation results involving factor scores to those of sum scores so that applied researchers can understand the advantages. Yet, researchers do not often compare sum scores and factor scores in terms of bias, a common simulation outcome. A reason for this is that sum scores are on a different scale and it is unclear how to compare sum scores to other types of scores. The purpose of this paper is to provide guidance for methodological researchers who wish to conduct research on scoring how to compute bias for sum scores by obtaining the expected values of their model parameters under a sum score model.

Measuring the intensity of emotions

We describe a theoretical framework for the measurement of the intensity of emotional experiences and summarize findings of a series of studies that implemented this framework. Our approach is based on a realist view of quantities and combines the modern psychometric (i.e., latent-variable) view of measurement with a deductive order of inquiry for testing measurement axioms. At the core of the method are nonmetric probabilistic difference scaling methods, a class of indirect scaling methods based on ordinal judgments of intensity differences. Originally developed to scale sensations and preferences, these scaling methods are also well-suited for measuring emotion intensity, particularly in basic research. They are easy to perform and provide scale values of emotion intensity that are much more precise than the typically used, quality-intensity emotion rating scales. Furthermore, the scale values appear to fulfill central measurement-theoretical axioms necessary for interval-level measurement. Because of these properties, difference scaling methods allow precise tests of emotion theories on the individual subject level.

Non-parametric Regression Among Factor Scores: Motivation and Diagnostics for Nonlinear Structural Equation Models

We provide a framework for motivating and diagnosing the functional form in the structural part of nonlinear or linear structural equation models when the measurement model is a correctly specified linear confirmatory factor model. A mathematical population-based analysis provides asymptotic identification results for conditional expectations of a coordinate of an endogenous latent variable given exogenous and possibly other endogenous latent variables, and theoretically well-founded estimates of this conditional expectation are suggested. Simulation studies show that these estimators behave well compared to presently available alternatives. Practically, we recommend the estimator using Bartlett factor scores as input to classical non-parametric regression methods.

The Mediating Role of Pain Cognitions and Pain Sensitivity in the Treatment Effect of Perioperative Pain Neuroscience Education in People Undergoing Surgery for Lumbar Radiculopathy

Though perioperative pain neuroscience education (PPNE) positively influences patients' surgical outcomes, little is known about the mechanisms behind this treatment's success. Therefore, this study aims to evaluate the potential mediating role of pain cognitions and pain sensitivity in the treatment effect of PPNE on postoperative quality of life in people undergoing surgery for lumbar radiculopathy. This secondary analysis uses data from 120 participants of a randomized controlled trial who were randomized to receive either PPNE or perioperative biomedical education before undergoing surgery for lumbar radiculopathy. Quality of life was assessed 1-year postsurgery using the short form 36-item health survey (SF36) physical and mental component scores. Potential mediators included pain cognitions (ie, kinesiophobia, pain catastrophizing, and hypervigilance) and pain sensitivity (ie, endogenous nociceptive modulation), assessed 6 weeks postsurgery. Mediation models were constructed using structural equation modeling, and 95% confidence intervals (CIs) were calculated using 10,000 bootstrap samples. Analyses show a significant total effect for PPNE (estimate = .464, 95% CI [.105, .825]) and a significant indirect effect via pain catastrophizing on the SF36 physical component (estimate = .124, 95% CI [.001, .293]). No mediating effect was found through the remaining pain cognitions or pain sensitivity measures. Also, no potential mediators were identified for the treatment effect of PPNE on the SF36 mental component. Our findings suggest that pain catastrophizing mediates the treatment effect of PPNE on physical health-related quality of life in people undergoing surgery for lumbar radiculopathy. PERSPECTIVE: This secondary analysis identified pain catastrophizing as a mediator for PPNE in people undergoing surgery for lumbar radiculopathy. More so, its findings indicate that this educational intervention can enhance the postoperative physical health-related quality of life of these patients by addressing their catastrophizing thoughts. TRIAL REGISTRATION: Clinicaltrials.gov (NCT02630732).

Investigating Moderation Effects at the Within-Person Level Using Intensive Longitudinal Data: A Two-Level Dynamic Structural Equation Modelling Approach in Mplus

Recent technological advances have provided new opportunities for the collection of intensive longitudinal data. Using methods such as dynamic structural equation modeling, these data can provide new insights into moment-to-moment dynamics of psychological and behavioral processes. In intensive longitudinal data (t > 20), researchers often have theories that imply that factors that change from moment to moment within individuals act as moderators. For instance, a person's level of sleep deprivation may affect how much an external stressor affects mood. Here, we describe how researchers can implement, test, and interpret dynamically changing within-person moderation effects using two-level dynamic structural equation modeling as implemented in the structural equation modeling software Mplus. We illustrate the analysis of within-person moderation effects using an empirical example investigating whether changes in spending time online using social media affect the moment-to-moment effect of loneliness on depressive symptoms, and highlight avenues for future methodological development. We provide annotated Mplus code, enabling researchers to better isolate, estimate, and interpret the complexities of within-person interaction effects.

More powerful parameter tests? No, rather biased parameter estimates. Some reflections on path analysis with weighted composites

Recently, a study compared the effect size and statistical power of covariance-based structural equation modeling (CB-SEM) and path analysis using various types of composite scores (Deng, L., & Yuan, K.-H., Behavior Research Methods, 55, 1460-1479, 2023). This comparison uses nine empirical datasets to estimate eleven models. Based on the meta-comparison, that study concludes that path analysis via weighted composites yields "path coefficients with less relative errors, as reflected by greater effect size and statistical power" (ibidem, p. 1475). In our paper, we object to this central conclusion. We demonstrate that the justification these authors provided for comparing CB-SEM and path analysis via weighted composites is not well grounded. Similarly, we explain that their employed study design, i.e., a meta-comparison, is very limited in its ability to compare the effect size and power delivered across these methods. Finally, we replicated Deng and Yuan's (ibidem) meta-comparison and show that CB-SEM using the normal-distribution-based maximum likelihood estimator does not necessarily deliver smaller effect sizes than path analysis via composites if a different scaling method is employed for CB-SEM.

A Cautionary Note Regarding Multilevel Factor Score Estimates from Lavaan

To compute factor score estimates, lavaan version 0.6–12 offers the function lavPredict( ) that can not only be applied in single-level modeling but also in multilevel modeling, where characteristics of higher-level units such as working environments or team leaders are often assessed by ratings of employees. Surprisingly, the function provides results that deviate from the expected ones. Specifically, whereas the function yields correct EAP estimates of higher-level factors, the ML estimates are counterintuitive and possibly incorrect. Moreover, the function does not provide the expected standard errors. I illustrate these issues using an example from organizational research where team leaders are evaluated by their employees, and I discuss these issues from a measurement perspective.

Changes in the Serum Concentration Levels of Serotonin, Tryptophan and Cortisol among Stress-Resilient and Stress-Susceptible Individuals after Experiencing Traumatic Stress

BACKGROUND

Stress is a common response to many environmental adversities. However, once dysregulated, this reaction can lead to psychiatric illnesses, such as post-traumatic stress disorder (PTSD). Individuals can develop PTSD after exposure to traumatic events, severely affecting their quality of life. Nevertheless, not all individuals exposed to stress will develop psychiatric disorders, provided they show enhanced stress-resilience mechanisms that enable them to successfully adapt to stressful situations and thus avoid developing a persistent psychopathology.

METHODS

The study involved 93 participants. Of them, 62 comprised a study group and 31 comprised a control group. The aim of the study was to assess serotonin, cortisol and tryptophan concentration levels in subjects with PTSD (stress-susceptible; PTSD-SS) and in healthy individuals (stress-resilient; PTSD-SR), who had experienced a traumatic event but fully recovered after the trauma. The subjects were between 18 and 50 years of age (mean 35.56 ± 8.26 years). The serum concentration levels of serotonin, cortisol and tryptophan were measured with an ELISA kit.

RESULTS

It was found that the serotonin, tryptophan and cortisol concentration levels were consistent with the features of both PTSD-SR and PTSD-SS patients. It was reported that the mean cortisol concentration levels increased more significantly in the PTSD-SS group than in the PTSD-SR group, versus those in the control group. Similarly, the PTSD-SS group was found to show a larger decrease in the mean serotonin concentration levels than the PTSD-SR group, versus those in the control group. No significant changes were found in the tryptophan concentration levels between the study groups, versus those in the control group.

CONCLUSIONS

These findings can be useful when attempting to improve resilience in individuals using neuropharmacological methods. However, it is necessary to conduct more cross-sectional studies that would address different types of negative stress to find out whether they share common pathways.