Charles C. Driver
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Charles C. Driver has a total of 53 co-authors. Here are the top co-authors by number of shared publications:
Manuel C. Voelkle
Psychological Research Methods, Humboldt University
6 shared publications- Christopher James Hopwood
Department of Psychology, University of Zurich
3 shared publications - Manuel C. Voelkle
Humboldt-Universität zu Berlin
3 shared publications Martin J. Tomasik
3 shared publications- Tobias Koch
Friedrich-Schiller-University Jena
3 shared publications - Johan H. L. Oud
Radboud University Nijmegen
2 shared publications
Latest Publications
- Open Access
- 24 August 2026
- Published by Center for Open Science in PsyArXiv
Sampling Fast and Slow Changes Inferences About Well-Being Dynamics
Subjective well-being unfolds over time, but inferences about its dynamics may depend on how often it is measured. We examined whether the same well-being processes yield different conclusions when observed annually versus bimonthly. Using two large German population-based panels, the annual SOEP study and the bimonthly GESIS panel, we fit continuous-time structural equation models to 10 cognitive and affective well-being domains. We used propensity-score matching to reduce observed demographic differences between panels. Our results showed that annual and bimonthly data were not fully exchangeable. Independently fitted models best captured the full predictive distribution. However, pooled models often predicted expected well-being levels as well as, or better than, independent models. Descriptive comparisons showed that baseline levels of subjective well-being were relatively similar across designs, whereas dynamic features differed. Deviations from baseline persisted four to eight times longer in annual than bimonthly data; long-run within-subject correlations were stronger in the bimonthly data for most domain pairs; and long-run within-subject variances were consistently larger in the annual data. Within-sample thinning analyses showed that sparser sampling did not necessarily alter recovered dynamics within the temporal ranges examined. These findings suggest that subjective well-being is a multi-timescale system where different observation schedules reveal different aspects of persistence, fluctuation, covariation, and heterogeneity. Temporal design should therefore be treated as a central determinant of substantive inferences.
- 24 August 2026
- Published by Center for Open Science in PsyArXiv
Open AccessSampling Fast and Slow Changes Inferences About Well-Being Dynamics
Subjective well-being unfolds over time, but inferences about its dynamics may depend on how often it is measured. We examined whether the same well-being processes yield different conclusions when observed annually versus bimonthly. Using two large German population-based panels, the annual SOEP study and the bimonthly GESIS panel, we fit continuous-time structural equation models to 10 cognitive and affective well-being domains. We used propensity-score matching to reduce observed demographic differences between panels. Our results showed that annual and bimonthly data were not fully exchangeable. Independently fitted models best captured the full predictive distribution. However, pooled models often predicted expected well-being levels as well as, or better than, independent models. Descriptive comparisons showed that baseline levels of subjective well-being were relatively similar across designs, whereas dynamic features differed. Deviations from baseline persisted four to eight times longer in annual than bimonthly data; long-run within-subject correlations were stronger in the bimonthly data for most domain pairs; and long-run within-subject variances were consistently larger in the annual data. Within-sample thinning analyses showed that sparser sampling did not necessarily alter recovered dynamics within the temporal ranges examined. These findings suggest that subjective well-being is a multi-timescale system where different observation schedules reveal different aspects of persistence, fluctuation, covariation, and heterogeneity. Temporal design should therefore be treated as a central determinant of substantive inferences.
- Open Access
- 21 July 2026
- Published by Center for Open Science in PsyArXiv
A Tutorial in Dynamic Systems Simulation and Modelling With R, ctsem, and lme4: Couples' Affect Dynamics Over Time
Understanding how people change is a fundamental goal of psychological science. However, translating complex ideas about psychological dynamics into formal models can be challenging without the right tools. In this tutorial, we introduce a approaches using R and the ctsem package to help researchers build and understand dynamic systems models that capture some of the complexity of psychological processes. We emphasize iterative model-building through simulations, model fitting, and visualizations of the model-implied dynamics alongside their fit to data. We begin with familiar linear models in lme4 and gradually transition to ctsem, which allows us to incorporate complexities such as state-dependent change, random fluctuation distinct from measurement error, covariate effects, interactions, and external inputs. These modelling concepts are illustrated using a running example of affect dynamics in couples therapy, demonstrating how key conceptual and methodological ideas in dynamic systems modelling come together. Our aim is to provide a general framework for understanding dynamic systems modelling and to encourage further exploration of theory-driven statistical approaches in psychological research.
- 21 July 2026
- Published by Center for Open Science in PsyArXiv
Open AccessA Tutorial in Dynamic Systems Simulation and Modelling With R, ctsem, and lme4: Couples' Affect Dynamics Over Time
Understanding how people change is a fundamental goal of psychological science. However, translating complex ideas about psychological dynamics into formal models can be challenging without the right tools. In this tutorial, we introduce a approaches using R and the ctsem package to help researchers build and understand dynamic systems models that capture some of the complexity of psychological processes. We emphasize iterative model-building through simulations, model fitting, and visualizations of the model-implied dynamics alongside their fit to data. We begin with familiar linear models in lme4 and gradually transition to ctsem, which allows us to incorporate complexities such as state-dependent change, random fluctuation distinct from measurement error, covariate effects, interactions, and external inputs. These modelling concepts are illustrated using a running example of affect dynamics in couples therapy, demonstrating how key conceptual and methodological ideas in dynamic systems modelling come together. Our aim is to provide a general framework for understanding dynamic systems modelling and to encourage further exploration of theory-driven statistical approaches in psychological research.
- Open Access
- 16 January 2026
- Published by Center for Open Science in PsyArXiv
Prediction of Performance in Standardised Assessments from Computer-Based Formative Assessment Data
Summative assessments (SAs) and formative assessments (FAs) fulfil complementary functions in the educational endeavour. SAs measure knowledge at the end of a unit in a standardised, high-stakes setting, while FAs evaluate student performance during daily classroom activities to tailor feedback and instruction. Computer-based FA (CBFA) systems enable collecting unprecedented amounts of data objectively and with minimal disruption for students, under conditions that more closely resemble real-life behaviour. Given concerns about student stress and ecological validity associated with SAs, potential biases in teacher judgements, and the high burden entailed by traditional classroom assessments, we investigated whether and how well FA outcomes can predict SA outcomes. Specifically, we estimated student abilities in a large sample of children evaluated at different time points during compulsory schooling and performed a systematic comparison of regression models trained to predict SA abilities on different subsets of features derived from FA abilities and auxiliary variables. A model that included mean abilities in different competence domains performed best, accounting for a considerable proportion of variance (30 - 48 %), although this was still below that explained by past SA measures. Most predictive FA features generally corresponded to abilities from the same or a similar competence domain as the predicted SA ability. We report systematic model biases that would warrant consideration when using the models for decision-making. Our findings provide valuable insights into how learning progress connects to future achievement, which can help teachers adapt instruction earlier and inform policies to reduce reliance on high-stakes testing.
- 16 January 2026
- Published by Center for Open Science in PsyArXiv
Open AccessPrediction of Performance in Standardised Assessments from Computer-Based Formative Assessment Data
Summative assessments (SAs) and formative assessments (FAs) fulfil complementary functions in the educational endeavour. SAs measure knowledge at the end of a unit in a standardised, high-stakes setting, while FAs evaluate student performance during daily classroom activities to tailor feedback and instruction. Computer-based FA (CBFA) systems enable collecting unprecedented amounts of data objectively and with minimal disruption for students, under conditions that more closely resemble real-life behaviour. Given concerns about student stress and ecological validity associated with SAs, potential biases in teacher judgements, and the high burden entailed by traditional classroom assessments, we investigated whether and how well FA outcomes can predict SA outcomes. Specifically, we estimated student abilities in a large sample of children evaluated at different time points during compulsory schooling and performed a systematic comparison of regression models trained to predict SA abilities on different subsets of features derived from FA abilities and auxiliary variables. A model that included mean abilities in different competence domains performed best, accounting for a considerable proportion of variance (30 - 48 %), although this was still below that explained by past SA measures. Most predictive FA features generally corresponded to abilities from the same or a similar competence domain as the predicted SA ability. We report systematic model biases that would warrant consideration when using the models for decision-making. Our findings provide valuable insights into how learning progress connects to future achievement, which can help teachers adapt instruction earlier and inform policies to reduce reliance on high-stakes testing.
- Open Access
- 6 October 2025
- Published by Center for Open Science in PsyArXiv
Score-Based Tests with Fixed Effects Person Parameters in Item Response Theory: Detecting Model Misspecification Including Differential Item Functioning
We present a fast, score-based test to detecting model misspecification in item response theory (IRT) models, that remains valid when person parameters are treated as fixed effects, as may be used for very large data sets. The new approximation (i) eliminates the need to pre-specify ability groups or priors for person abilities, (ii) does not require explicit functional form assumptions, (iii) works with two estimators designed for very high item / person counts -- constrained joint maximum likelihood (CJML) and joint maximum a posteriori (JMAP) -- and (iv) requires only a single model fit, making DIF-screening faster and simpler than alternatives based on model comparisons. A spline-based residualisation step further suppresses spurious Type I error when the ordering covariate is correlated with ability. Simulations with the two-parameter logistic model show nominal error rates and high power once examinees contribute 15–20 responses; only extremely short tests (10 items) still pose challenges under strong impact. An application to 1602 reading items and 57684 students from the Mindsteps platform demonstrates scalability and practical value, flagging 13% of items for gender-related DIF and correlating highly with conventional approaches of explicitly modelling DIF. Together, these results position the proposed test as a robust, computation-light diagnostic for large-scale assessments when classical random-effects approaches are infeasible, ability group structure is unknown or complex, or the shape of DIF effects is unknown or complex.
- 6 October 2025
- Published by Center for Open Science in PsyArXiv
Open AccessScore-Based Tests with Fixed Effects Person Parameters in Item Response Theory: Detecting Model Misspecification Including Differential Item Functioning
We present a fast, score-based test to detecting model misspecification in item response theory (IRT) models, that remains valid when person parameters are treated as fixed effects, as may be used for very large data sets. The new approximation (i) eliminates the need to pre-specify ability groups or priors for person abilities, (ii) does not require explicit functional form assumptions, (iii) works with two estimators designed for very high item / person counts -- constrained joint maximum likelihood (CJML) and joint maximum a posteriori (JMAP) -- and (iv) requires only a single model fit, making DIF-screening faster and simpler than alternatives based on model comparisons. A spline-based residualisation step further suppresses spurious Type I error when the ordering covariate is correlated with ability. Simulations with the two-parameter logistic model show nominal error rates and high power once examinees contribute 15–20 responses; only extremely short tests (10 items) still pose challenges under strong impact. An application to 1602 reading items and 57684 students from the Mindsteps platform demonstrates scalability and practical value, flagging 13% of items for gender-related DIF and correlating highly with conventional approaches of explicitly modelling DIF. Together, these results position the proposed test as a robust, computation-light diagnostic for large-scale assessments when classical random-effects approaches are infeasible, ability group structure is unknown or complex, or the shape of DIF effects is unknown or complex.
- Open Access
- 16 September 2025
- Published by Center for Open Science in PsyArXiv
Personality functioning as generalized correlated changes in personality traits
Contemporary personality disorder diagnosis parses personality functioning (PF) as a general criterion for diagnosis and personality traits as the style in which the disorder manifests. There is ongoing debate about the distinctness of PF and personality traits, but research on core differences between PF and personality traits has been limited by an over-reliance on cross-sectional data that may be insensitive to these differences. We fit a continuous time model to five factor model (FFM) personality trait facets that parsed general correlated change and change that could be accounted for by FFM domains using ten years of longitudinal data from the Collaborative Longitudinal Personality disorders Study (N = 733). We found that the general change factor was very similar to lay and expert conceptions of psychological health and explained more variance in personality disorder symptom reductions over time than trait change factors. These results support the distinction between PF and personality traits, suggest that features of PF can be identified within change processes in personality trait assessments, and highlight the importance of designs that are sensitive to theoretical differences in these concepts.
- 16 September 2025
- Published by Center for Open Science in PsyArXiv
Open AccessPersonality functioning as generalized correlated changes in personality traits
Contemporary personality disorder diagnosis parses personality functioning (PF) as a general criterion for diagnosis and personality traits as the style in which the disorder manifests. There is ongoing debate about the distinctness of PF and personality traits, but research on core differences between PF and personality traits has been limited by an over-reliance on cross-sectional data that may be insensitive to these differences. We fit a continuous time model to five factor model (FFM) personality trait facets that parsed general correlated change and change that could be accounted for by FFM domains using ten years of longitudinal data from the Collaborative Longitudinal Personality disorders Study (N = 733). We found that the general change factor was very similar to lay and expert conceptions of psychological health and explained more variance in personality disorder symptom reductions over time than trait change factors. These results support the distinction between PF and personality traits, suggest that features of PF can be identified within change processes in personality trait assessments, and highlight the importance of designs that are sensitive to theoretical differences in these concepts.
- Open Access
- 9 September 2025
- Published by Center for Open Science in PsyArXiv
Coupled Cognitive and Social Development From Childhood To Adolescence: Behavioral Problems Linked To Lower Levels and Less Social-Cognitive Integration
Cognitive skills tend to grow and change in interdependent ways across development, yet it remains unclear whether (a) early patterns of cognitive coupling depend on social-skill development and (b) whether individual differences in such coupling have predictive value for later behavioral outcomes. We addressed these questions using eight repeated assessments from the NICHD–SECCYD longitudinal study, tracking children from early childhood through adolescence. Using continuous-time structural equation models, we modelled dynamic, reciprocal interactions among cognitive (e.g., vocabulary, attention, memory, executive functioning) and social (e.g., assertion, cooperation, responsibility, self-control, peer competence) domains from age 4.5 to 15 years. Cognitive traits showed stronger overall growth and rank-order stability than social traits, whereas social-emotional skills displayed tighter internal coupling. Executive functioning emerged as a central connector across domains, showing robust cross-domain associations with multiple social skills. Children who exhibited more behavioral problems at age 15 had weaker early cross-domain trait correlations—particularly between (rather than within) cognitive and social domains—indicating lower early developmental integration. By adolescence, these children showed increased cohesion within the cognitive domain but reduced coupling with social traits, suggesting a shift toward domain-specific rather than integrated development. These findings highlight how early differences in the co-development of cognitive and social abilities may shape later behavioral adjustment, supporting a dynamic, systems-level view of developmental risk and adaptation.
- 9 September 2025
- Published by Center for Open Science in PsyArXiv
Open AccessCoupled Cognitive and Social Development From Childhood To Adolescence: Behavioral Problems Linked To Lower Levels and Less Social-Cognitive Integration
Cognitive skills tend to grow and change in interdependent ways across development, yet it remains unclear whether (a) early patterns of cognitive coupling depend on social-skill development and (b) whether individual differences in such coupling have predictive value for later behavioral outcomes. We addressed these questions using eight repeated assessments from the NICHD–SECCYD longitudinal study, tracking children from early childhood through adolescence. Using continuous-time structural equation models, we modelled dynamic, reciprocal interactions among cognitive (e.g., vocabulary, attention, memory, executive functioning) and social (e.g., assertion, cooperation, responsibility, self-control, peer competence) domains from age 4.5 to 15 years. Cognitive traits showed stronger overall growth and rank-order stability than social traits, whereas social-emotional skills displayed tighter internal coupling. Executive functioning emerged as a central connector across domains, showing robust cross-domain associations with multiple social skills. Children who exhibited more behavioral problems at age 15 had weaker early cross-domain trait correlations—particularly between (rather than within) cognitive and social domains—indicating lower early developmental integration. By adolescence, these children showed increased cohesion within the cognitive domain but reduced coupling with social traits, suggesting a shift toward domain-specific rather than integrated development. These findings highlight how early differences in the co-development of cognitive and social abilities may shape later behavioral adjustment, supporting a dynamic, systems-level view of developmental risk and adaptation.
- 1 April 2025
- Published by American Psychological Association (APA) in Journal of Personality and Social Psychology
- Vol. 128 (4) , 930-948
The dynamics of self-esteem and depressive symptoms across days, months, and years.
Self-esteem and depressive symptoms are important predictors of a range of societally relevant outcomes and are theorized to influence each other reciprocally over time. However, existing research offers only a limited understanding of how their dynamics unfold across different timescales. Using three data sets with different temporal resolutions, we aimed to advance our understanding of the temporal unfolding of the reciprocal dynamics between self-esteem and depressive symptoms. Across these data sets, participants (Ntotal = 6,210) rated their self-esteem and depressive symptoms between 6 and 14 times across days, months, and years, respectively. Using continuous time dynamic models, we found limited evidence for significant within-person cross-lagged effects between self-esteem and depressive symptoms. Only in the yearly data set, a cross-lagged effect from depressive symptoms to self-esteem emerged quite consistently. However, in all data sets, cross-lagged effects were small in size (-0.04 ≤ β ≤ -0.01). These findings suggest that the reciprocal dynamics between self-esteem and depressive symptoms may be less robust than commonly thought. Furthermore, exploratory analyses indicated that these effects depended on people's overall levels of depressive symptoms, suggesting that theoretical frameworks that highlight transactions between self-esteem and depression may not generalize across all levels of depressive symptoms. Finally, self-esteem and depressive symptoms were strongly correlated within measurements, similarly stable over time, and changed similarly in response to negative life events, provoking questions as to their conceptual distinctiveness and measurement approaches. (PsycInfo Database Record (c) 2025 APA, all rights reserved).
- 1 April 2025
- Published by American Psychological Association (APA) in Journal of Personality and Social Psychology
- Vol. 128 (4) , 930-948
The dynamics of self-esteem and depressive symptoms across days, months, and years.
Self-esteem and depressive symptoms are important predictors of a range of societally relevant outcomes and are theorized to influence each other reciprocally over time. However, existing research offers only a limited understanding of how their dynamics unfold across different timescales. Using three data sets with different temporal resolutions, we aimed to advance our understanding of the temporal unfolding of the reciprocal dynamics between self-esteem and depressive symptoms. Across these data sets, participants (Ntotal = 6,210) rated their self-esteem and depressive symptoms between 6 and 14 times across days, months, and years, respectively. Using continuous time dynamic models, we found limited evidence for significant within-person cross-lagged effects between self-esteem and depressive symptoms. Only in the yearly data set, a cross-lagged effect from depressive symptoms to self-esteem emerged quite consistently. However, in all data sets, cross-lagged effects were small in size (-0.04 ≤ β ≤ -0.01). These findings suggest that the reciprocal dynamics between self-esteem and depressive symptoms may be less robust than commonly thought. Furthermore, exploratory analyses indicated that these effects depended on people's overall levels of depressive symptoms, suggesting that theoretical frameworks that highlight transactions between self-esteem and depression may not generalize across all levels of depressive symptoms. Finally, self-esteem and depressive symptoms were strongly correlated within measurements, similarly stable over time, and changed similarly in response to negative life events, provoking questions as to their conceptual distinctiveness and measurement approaches. (PsycInfo Database Record (c) 2025 APA, all rights reserved).
- Open Access
- 28 February 2025
- Published by Center for Open Science in PsyArXiv
The Dynamics of Self-Esteem and Depressive Symptoms Across Days, Months, and Years
Self-esteem and depressive symptoms are important predictors of a range of societally relevant outcomes and are theorized to influence each other reciprocally over time. However, existing research offers only a limited understanding of how their dynamics unfold across different time scales. Using three datasets with different temporal resolutions, we aimed to advance our understanding of the temporal unfolding of the reciprocal dynamics between self-esteem and depressive symptoms. Across these datasets, participants (Ntotal = 6,210) rated their self-esteem and depressive symptoms between 6 and 14 times across days, months, and years, respectively. Using continuous time dynamic models, we found limited evidence for significant within-person cross-lagged effects between self-esteem and depressive symptoms. Only in the yearly dataset, a cross-lagged effect from depressive symptoms to self-esteem emerged quite consistently. However, in all datasets, cross-lagged effects were small in size (–0.04 ≤ ≤ –0.01). These findings suggest that the reciprocal dynamics between self-esteem and depressive symptoms may be less robust than commonly thought. Furthermore, exploratory analyses indicated that these effects depended on people’s overall levels of depressive symptoms, suggesting that theoretical frameworks that highlight transactions between self-esteem and depression may not generalize across all levels of depressive symptoms. Finally, self-esteem and depressive symptoms were strongly correlated within measurements, similarly stable over time, and changed similarly in response to negative life events, provoking questions as to their conceptual distinctiveness and measurement approaches.
- 28 February 2025
- Published by Center for Open Science in PsyArXiv
Open AccessThe Dynamics of Self-Esteem and Depressive Symptoms Across Days, Months, and Years
Self-esteem and depressive symptoms are important predictors of a range of societally relevant outcomes and are theorized to influence each other reciprocally over time. However, existing research offers only a limited understanding of how their dynamics unfold across different time scales. Using three datasets with different temporal resolutions, we aimed to advance our understanding of the temporal unfolding of the reciprocal dynamics between self-esteem and depressive symptoms. Across these datasets, participants (Ntotal = 6,210) rated their self-esteem and depressive symptoms between 6 and 14 times across days, months, and years, respectively. Using continuous time dynamic models, we found limited evidence for significant within-person cross-lagged effects between self-esteem and depressive symptoms. Only in the yearly dataset, a cross-lagged effect from depressive symptoms to self-esteem emerged quite consistently. However, in all datasets, cross-lagged effects were small in size (–0.04 ≤ ≤ –0.01). These findings suggest that the reciprocal dynamics between self-esteem and depressive symptoms may be less robust than commonly thought. Furthermore, exploratory analyses indicated that these effects depended on people’s overall levels of depressive symptoms, suggesting that theoretical frameworks that highlight transactions between self-esteem and depression may not generalize across all levels of depressive symptoms. Finally, self-esteem and depressive symptoms were strongly correlated within measurements, similarly stable over time, and changed similarly in response to negative life events, provoking questions as to their conceptual distinctiveness and measurement approaches.
- 1 February 2025
- Published by American Psychological Association (APA) in Psychological Methods
- Vol. 30 (1) , 174-202
Inference with cross-lagged effects—Problems in time.
The interpretation of cross-effects from vector autoregressive models to infer structure and causality among constructs is widespread and sometimes problematic. I describe problems in the interpretation of cross-effects when processes that are thought to fluctuate continuously in time are, as is typically done, modeled as changing only in discrete steps (as in e.g., structural equation modeling)-zeroes in a discrete-time temporal matrix do not necessarily correspond to zero effects in the underlying continuous processes, and vice versa. This has implications for the common case when the presence or absence of cross-effects is used for inference about underlying causal processes. I demonstrate these problems via simulation, and also show that when an underlying set of processes are continuous in time, even relatively few direct causal links can result in much denser temporal effect matrices in discrete-time. I demonstrate one solution to these issues, namely parameterizing the system as a stochastic differential equation and focusing inference on the continuous-time temporal effects. I follow this with some discussion of issues regarding the switch to continuous-time, specifically regularization, appropriate measurement time lag, and model order. An empirical example using intensive longitudinal data highlights some of the complexities of applying such approaches to real data, particularly with respect to model specification, examining misspecification, and parameter interpretation. (PsycInfo Database Record (c) 2024 APA, all rights reserved).
- 1 February 2025
- Published by American Psychological Association (APA) in Psychological Methods
- Vol. 30 (1) , 174-202
Inference with cross-lagged effects—Problems in time.
The interpretation of cross-effects from vector autoregressive models to infer structure and causality among constructs is widespread and sometimes problematic. I describe problems in the interpretation of cross-effects when processes that are thought to fluctuate continuously in time are, as is typically done, modeled as changing only in discrete steps (as in e.g., structural equation modeling)-zeroes in a discrete-time temporal matrix do not necessarily correspond to zero effects in the underlying continuous processes, and vice versa. This has implications for the common case when the presence or absence of cross-effects is used for inference about underlying causal processes. I demonstrate these problems via simulation, and also show that when an underlying set of processes are continuous in time, even relatively few direct causal links can result in much denser temporal effect matrices in discrete-time. I demonstrate one solution to these issues, namely parameterizing the system as a stochastic differential equation and focusing inference on the continuous-time temporal effects. I follow this with some discussion of issues regarding the switch to continuous-time, specifically regularization, appropriate measurement time lag, and model order. An empirical example using intensive longitudinal data highlights some of the complexities of applying such approaches to real data, particularly with respect to model specification, examining misspecification, and parameter interpretation. (PsycInfo Database Record (c) 2024 APA, all rights reserved).