Postdoctoral Researcher @ UZH

Pascal Küng.

Investigating dyadic data, intensive longitudinal designs, and the dynamics of daily health behavior change.

CONSOLE // wbCorractive_session

>install.packages("wbCorr")

R Packagev0.1.0

wbCorr Package

Clean estimation of within-person and between-person correlations in repeated-measures datasets, decomposing variances across multilevel structures.

View on GitHub→

DYADIC DYNAMICSin_sync

Support Reactance

Theory Focus

Social Control in Couples

Investigating how partners regulate daily health behaviors, and how individual preferences for autonomy influence support dynamics and emotional reactance.

Explore Theory→

VISUALIZER // MULTILEVELcompiled

X: Independent VariableY: Outcome

Teaching

Methods & Demos

Developing visual intuition for advanced concepts like random slopes, APIM, and DIM equivalence through highly interactive web demonstrators.

Try Demos→

01 —

Research Focus

Methodological rigour
meeting empirical questions.

Modeling interpersonal processes and relational dynamics in close relationships. I investigate how partners mutually influence each other's health behaviors over time.

APIMActor-Partner EffectsInterdependenceCouples Research

Hover over nodes, arrows, or paths in the APIM diagram to explore relational effects and interpersonal influences.

Capturing life as it is lived. Using Ecological Momentary Assessment (EMA) and daily diaries to study within-person micro-processes and health behavior changes in real time.

EMA / Daily DiariesWithin-Person dynamicsMultilevel ModelingTime-Series

Hover over the chart to scan daily assessments, or hover the labels to see variance decomposition (within vs. between-person).

Bridging the gap between conceptual models and statistical evidence. Translating psychological theories of social control and autonomy into precise, interpretable empirical tests.

MeasurementInterpretationConstruct ValidityBayesian Models

Hover over components of the measurement model to see how abstract theory is linked to concrete empirical instruments.

02 —

Highlighted Research

Interdependent processes in daily health behavior change.

Core Topic

Social Control in Couples

Investigating how health-related social control interacts with individual factors like the preference for self-reliance. Using daily diary data and dyadic EMA designs to understand behavioral trajectories and affective reactance.

View social control publications →

R Package

wbCorr Package

A dedicated methodological package designed to cleanly estimate bivariate within-person and between-person correlations in clustered, repeated-measures datasets.

View on GitHub

03 —

Scholarly Directory

Publications & research outputs.

2026First Author

Does preference for self-reliance moderate associations of health-related social control with physical activity and smoking cessation? Two intensive longitudinal studies.

Read Paper→

Küng, P., Bierbauer, W., Berli, C., Höhener, P. S., Lüscher, J., Bermudez, T., Banik, A., Łuszczyńska, A., & Scholz, U.

British Journal of Health Psychology•31(2), e70082

#IntensiveLongitudinal#SocialControl#Moderation#PhysicalActivity

2026First Author

Daily associations of three types of social control with physical activity, reactance, and mood in romantic couples: A dyadic intensive longitudinal study.

Read Paper→

Küng, P., Höhener, P. S., Allen, J. M., Tobias, R., & Scholz, U.

Psychology & Health•1–25

#DyadicData#EMA/DailyDiary#Reactance#PhysicalActivity

2025First Author

Health-related social control in overweight romantic couples: Daily associations with physical activity and affect for targets and agents.

Read Paper→

Küng, P., Berli, C., Höhener, P. S., Tobias, R., & Scholz, U.

Annals of Behavioral Medicine•59(1), kaae093

#SocialControl#DyadicData#EMA/DailyDiary#OverweightCouples

2025Software

Tutorial: Distinguishable and exchangeable dyads -- Bayesian multilevel modelling

Code / Link→

Küng, P.

Zenodo Tutorial•Version 2.0.1

#BayesianMLM#DyadicData#Tutorial#R/brms

2023Software

wbCorr: Bivariate within- and between-cluster correlations

Code / Link→

Küng, P.

R Package•Version 0.1.22

#RPackage#Methodology#Correlation#ClusteredData

2026Co-Authored

Effectiveness of smartphone-based dyadic interventions to increase physical activity in romantic couples: Microrandomized trial.

Read Paper→

Höhener, P. S., Tobias, R., Allen, J. M., Küng, P., & Scholz, U.

JMIR mHealth and uHealth•14, e67136

#mHealth#Micro-randomizedTrial#DyadicIntervention#PhysicalActivity

2025Co-Authored

Investigating the association between companionship and behavior concordance: A dyadic panel study.

Read Paper→

Höhener, P. S., Lüscher, J., Tobias, R., Küng, P., & Scholz, U.

Personal Relationships•32(2), e70005

#Companionship#Concordance#DyadicPanelStudy#Relationships

04 —

Methods Intuition

Visualizing multilevel models.

In intensive longitudinal and dyadic designs, mapping individual slopes or visualizing model reparameterizations makes abstract equations intuitive. Toggle the tabs below to explore random effects or check the mathematical equivalence of APIM and DIM.

Parameters

Fixed Intercept (β₀)7.0

SD Random Intercepts (σ_u₀)1.0

Fixed Slope (β₁)-0.25

SD Random Slopes (σ_u₁)0.25

Intercept-Slope Corr. (ρ)-0.40

Fixed Effects Line (Drag Handles)

10 Simulated Individuals (u_i)

Y_{ij} = ({\color{#4E94F8} 7.0} + {\color{#93C5FD} u_{0i}}) + ({\color{#E25C5C} -0.25} + {\color{#FCA5A5} u_{1i}})X_{ij}

{\color{#93C5FD} u_{0i}} \sim N(0, {\color{#93C5FD} 1.0}^2), \quad {\color{#FCA5A5} u_{1i}} \sim N(0, {\color{#FCA5A5} 0.25}^2)

\text{Cor}({\color{#93C5FD} u_{0i}}, {\color{#FCA5A5} u_{1i}}) = {\color{#C59E5D} -0.40}

💡 Hover over an individual's gray line to see their unique formula realization

💡 Drag plot nodes to change intercept & slope

Predictors

APIM Space (Individual)

Actor Predictor (

x_{\\text{actor}}

)0.0

Partner Predictor (

x_{\\text{partner}}

)0.0

DIM Space (Couples)

Between-Couple Mean (

x_{\\text{cbc}}

)0.0

Within-Couple Dev (

x_{\\text{cwc}}

)0.0

Coordinate Relations:

Between-Couple ( $x_{\\text{cbc}}$ ) = ( $x_{\\text{actor}}$ + $x_{\\text{partner}}$ ) / 2
Within-Couple ( $x_{\\text{cwc}}$ ) = ( $x_{\\text{actor}}$ - $x_{\\text{partner}}$ ) / 2

Actor-Partner model (APIM)

\widehat{Y}_{\text{APIM}} = 5.00 + 1.70 \times {\color{#4E94F8}(\phantom{-}0.00)} + 0.25 \times {\color{#A78BFA}(\phantom{-}0.00)} = \phantom{-}\phantom{0}5.00

Dyad-Individual model (DIM)

\widehat{Y}_{\text{DIM}} = 5.00 + 1.95 \times {\color{#E26A5C}(\phantom{-}0.00)} + 1.45 \times {\color{#FBBF24}(\phantom{-}0.00)} = \phantom{-}\phantom{0}5.00

b_{\text{cbc}} = b_{\text{actor}} + b_{\text{partner}} \implies 1.95 = 1.70 + 0.25

b_{\text{cwc}} = b_{\text{actor}} - b_{\text{partner}} \implies 1.45 = 1.70 - 0.25

Drag point inside the 2D grid to update coordinates

Why the sum & difference?

1. Between-Couple Effect (The Sum)

Suppose both partners increase their predictor score jointly by 1 unit (the actor and the partner go up by 1 unit, meaning the couple mean also goes up by 1 unit)(moving along the

x_{\\text{actor}} = x_{\\text{partner}}

diagonal):

In the individual-level model (APIM), this jointly adds $b_{\text{actor}} \times 1$ from the actor plus $b_{\text{partner}} \times 1$ from the partner.
In the couple-level model (DIM), this corresponds to a 1-unit increase in the couple mean ( ${\color{#E26A5C}x_{\text{cbc}}} = 1$ ) while the couple deviation remains zero ( ${\color{#FBBF24}x_{\text{cwc}}} = 0$ ).

Therefore, the couple-level mean effect ( $b_{\text{cbc}}$ ) must equal the sum of both individual effects:

b_{\text{cbc}} = b_{\text{actor}} + b_{\text{partner}} = 1.70 + 0.25 = 1.95

2. Within-Couple Effect (The Difference)

Suppose the actor increases by 1 unit and the partner decreases by 1 unit (meaning the couple mean remains unchanged, but the couple deviation goes up by 1 unit)(moving along the

x_{\\text{actor}} = -x_{\\text{partner}}

diagonal):

In the individual-level model (APIM), this adds $b_{\text{actor}} \times 1$ but subtracts $b_{\text{partner}} \times 1$ .
In the couple-level model (DIM), this corresponds to a 1-unit increase in the couple deviation ( ${\color{#FBBF24}x_{\text{cwc}}} = 1$ ) while the couple mean remains zero ( ${\color{#E26A5C}x_{\text{cbc}}} = 0$ ).

Therefore, the couple-level deviation effect ( $b_{\text{cwc}}$ ) must equal the difference between individual effects:

b_{\text{cwc}} = b_{\text{actor}} - b_{\text{partner}} = 1.70 - 0.25 = 1.45

05 —

Get in touch

Let's
connect.

Interested in collaboration, methodology advisory, or longitudinal research discussions? My inbox is always open.

Email Me →

GitHub ORCID

Google Scholar / Publications GitHub Repository

Pascal Küng.

wbCorr Package

Social Control in Couples

Methods & Demos

Methodological rigour meeting empirical questions.

Interdependent processes in daily health behavior change.

Social Control in Couples

wbCorr Package

Publications & research outputs.

Does preference for self-reliance moderate associations of health-related social control with physical activity and smoking cessation? Two intensive longitudinal studies.

Daily associations of three types of social control with physical activity, reactance, and mood in romantic couples: A dyadic intensive longitudinal study.

Health-related social control in overweight romantic couples: Daily associations with physical activity and affect for targets and agents.

Tutorial: Distinguishable and exchangeable dyads -- Bayesian multilevel modelling

wbCorr: Bivariate within- and between-cluster correlations

Effectiveness of smartphone-based dyadic interventions to increase physical activity in romantic couples: Microrandomized trial.

Investigating the association between companionship and behavior concordance: A dyadic panel study.

Visualizing multilevel models.

Parameters

Predictors

1. Between-Couple Effect (The Sum)

2. Within-Couple Effect (The Difference)

Let's connect.

Methodological rigour
meeting empirical questions.

Let's
connect.