Power Analysis for the Random Intercept Cross-Lagged Panel Model Using the powRICLPM R-Package

Figures & data

Figure 1. A bivariate random intercept cross-lagged panel model with three waves of data. α_t and δ_t are autoregressive effects of W_X and W_Y, respectively. γ_t and β_t are the cross-lagged effects of $W_{X,t-1}$ on $W_{Y,t}$ and $W_{Y,t-1}$ on $W_{X,t},$ respectively. The mean structure with ${\mu }_{X,t}$ and ${\mu }_{Y,t}$ is not explicitly included here.

Figure 2. Overview of power analysis procedure used by powRICLPM.

Figure 3. Results of preliminary power analysis for the RI-CLPM, based on 100 replications, for a cross-lagged effect of 0.10 (standardized). The different panels display results for conditions with a 0.3, 0.5 and 0.7 proportion of between-unit variance, respectively. The vertical error bars represent the uncertainty around the simulated power.

Figure 4. Results of the validation phase of the power analysis for the RI-CLPM, based on 2000 replications. The different panels display results for conditions with a 0.3, 0.5 and 0.7 proportion of between-unit variance, respectively. The vertical error bars represent the uncertainty around the simulated power.