Insight Into Individual Differences in Emotion Dynamics With Clustering

Ernst, Anja F.; Timmerman, Marieke E.; Jeronimus, Bertus F.; Albers, Casper J.

doi:10.1177/1073191119873714

Cited by 17 publications

(14 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is possible that different individuals account for covariance of different between-person paths (Moeller, 2021). New statistical methods are on the way, such as latent class vector-autoregressive modeling, in which individuals with similar affective time-series patterns can be identified (Ernst et al, 2021).…”

Section: Limitationsmentioning

confidence: 99%

Beyond main effects? Affect level as a moderator in the relation between affect dynamics and depressive symptoms

Maciejewski¹,

Roekel²,

Ha³

et al. 2022

Preprint

View full text Add to dashboard Cite

Higher affect variability, instability, and inertia in daily life are usually seen as indicators of emotional dysregulation. Research has shown that individuals with such affect dynamic patterns experience more depressive symptoms. However, similar affect dynamics might function differently across individuals. In this study, we propose that the impact of affect dynamics on depressive symptoms depends on how individuals feel on average (i.e., their affect levels). We analyzed data from seven studies that measured affect in daily life (N = 1,448, age range = 11.7-29.9 years, 64.8% females). Main and interaction effects of affect dynamics (variability, instability, inertia) and affect level on depressive symptoms were tested, separately for positive affect (PA) and negative affect (NA). For PA, we found mostly main effects of PA dynamics and PA level on depressive symptoms, but no interactions, indicating that PA dynamics are associated with depressive symptoms independent of how individuals feel on average. For NA, significant interactions emerged for NA variability × NA level, NA instability × NA level, but not NA inertia × NA level. For individuals with low NA levels, high NA variability and NA instability were associated with more depressive symptoms. In contrast, for individuals with high NA levels, high NA variability (but not NA instability) was associated with reduced depressive symptoms. These results indicate that high affect variability may not always signal emotional dysregulation and even be beneficial for those with high NA levels. Overall, this underscores the need for a more nuanced understanding of affect variability in depressive symptoms.

show abstract

Section: Limitationsmentioning

confidence: 99%

Beyond main effects? Affect level as a moderator in the relation between affect dynamics and depressive symptoms

Maciejewski¹,

Roekel²,

Ha³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…To briefly investigate estimation accuracy (see "Appendix") also, we did not vary the regressive parameters within each design cell. 2 The value of the auto-and cross-regressive effects are based on previous simulation studies (see Bulteel et al 2016a;Ernst et al 2019;Liu 2017) and empirical applications of these models to intensive longitudinal data (see Krone et al 2018;Kuppens et al 2010). Throughout the simulation settings, we set the standard deviation of the within-individual innovations σ to one for all individuals and the covariances to 0.2.…”

Section: Simulation Designmentioning

confidence: 99%

“…Therefore, for each particular variable, the model allows estimating the effect of its past values on current 433 values (i.e., autoregressive effects) as well as the effect of past values of the rest of the variables (i.e., cross-regressive effects). This model can be estimated at the individual level (i.e., personspecific VAR(1)) or using a multilevel (see, e.g., Asparouhov et al 2018;Bulteel et al 2018a;Bringmann et al 2013;Bringmann et al 2016;Jongerling et al 2015) or clustering (e.g., Bulteel et al 2016a;Ernst et al 2019) framework. Within the multilevel approach, the autoregressive and cross-regressive effects can be considered as random variables that vary across individuals.…”

Section: Introductionmentioning

confidence: 99%

“…As we indicated above, one can model the data of each person separately or use a multilevel approach. Moreover, a number of clustering techniques have been proposed as well (e.g., Bulteel et al 2016a;Ernst et al 2019), that can be useful in case one assumes the individual differences to be at least partly categorical. Here, the question becomes how to know which approach captures the individual differences well and how this is affected by the previously mentioned data char-predict future states.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Systematic Study into the Factors that Affect the Predictive Accuracy of Multilevel VAR(1) Models

2021

View full text Add to dashboard Cite

The use of multilevel VAR(1) models to unravel within-individual process dynamics is gaining momentum in psychological research. These models accommodate the structure of intensive longitudinal datasets in which repeated measurements are nested within individuals. They estimate within-individual auto-and cross-regressive relationships while incorporating and using information about the distributions of these effects across individuals. An important quality feature of the obtained estimates pertains to how well they generalize to unseen data. Bulteel and colleagues (Psychol Methods 23(4):740-756, 2018a) showed that this feature can be assessed through a cross-validation approach, yielding a predictive accuracy measure. In this article, we follow up on their results, by performing three simulation studies that allow to systematically study five factors that likely affect the predictive accuracy of multilevel VAR(1) models: (i) the number of measurement occasions per person, (ii) the number of persons, (iii) the number of variables, (iv) the contemporaneous collinearity between the variables, and (v) the distributional shape of the individual differences in the VAR(1) parameters (i.e., normal versus multimodal distributions). Simulation results show that pooling information across individuals and using multilevel techniques prevent overfitting. Also, we show that when variables are expected to show strong contemporaneous correlations, performing multilevel VAR(1) in a reduced variable space can be useful. Furthermore, results reveal that multilevel VAR(1) models with random effects have a better predictive performance than person-specific VAR(1) models when the sample includes groups of individuals that share similar dynamics.

show abstract

“…They used a non-parametric trajectory representation, allowing for high flexibility in the shape of the group trajectories. Lastly, Ernst et al (2019) analyzed ecological momentary assessments of subjects, assessing their emotional state three times per day over a period of 30 days. Subgroups with different emotion dynamics were discovered by clustering subjects based on individual vector autoregressive model coefficients.…”

Section: Introductionmentioning

confidence: 99%

A comparison of methods for clustering longitudinal data with slowly changing trends

Teuling

Pauws

Heuvel

2021

Communications in Statistics - Simulation and Computation

View full text Add to dashboard Cite

Longitudinal clustering provides a detailed yet comprehensible description of time profiles among subjects. With several approaches that are commonly used for this purpose, it remains unclear under which conditions a method is preferred over another method. We investigated the performance of five methods using Monte Carlo simulations on synthetic datasets, representing various scenarios involving polynomial time profiles. The performance was evaluated on two aspects: The agreement of the group assignment to the simulated reference, as measured by the split-join distance, and the trend estimation error, as measured by a weighted minimum of the mean squared error (WMMSE). Growth mixture modeling (GMM) was found to achieve the best overall performance, followed closely by a two-step approach using growth curve modeling and k-means (GCKM). Considering the model similarities between GMM and GCKM, the latter is preferred for large datasets for its computational efficiency. Longitudinal k-means (KML) and group-based trajectory modeling were found to have practically identical solutions in the case that the group trajectory model of the latter method is correctly specified. Both methods performed less than GMM and GCKM in most settings.

show abstract

Insight Into Individual Differences in Emotion Dynamics With Clustering

Cited by 17 publications

References 57 publications

Beyond main effects? Affect level as a moderator in the relation between affect dynamics and depressive symptoms

Beyond main effects? Affect level as a moderator in the relation between affect dynamics and depressive symptoms

A Systematic Study into the Factors that Affect the Predictive Accuracy of Multilevel VAR(1) Models

A comparison of methods for clustering longitudinal data with slowly changing trends

Contact Info

Product

Resources

About