TABLE 4.
Predictiveness of different factors (i.e., ml, g, rl, and d) regarding the criterion cogIll in two different frequentist models (with and without interactions).
Frequentist Model | Estimate | SE | p | Explained variance |
Without Interactions | R2Marginal = 0.054; R2Conditional = 0.040 | |||
γ00 | −2.2 | 0.575 | < 0.001 | |
ml | 0.067 | 0.006 | < 0.001 | |
g | 0.039 | 0.012 | 0.001 | |
rl | −0.028 | 0.014 | 0.04 | |
d | –1.074 | 0.804 | 0.182 | |
With Interactions | R2Marginal = 0.107; R2Conditional = 0.348 | |||
γ00 | −3.039 | 0.615 | < 0.001 | |
ml | 0.087 | 0.008 | < 0.001 | |
g | 0.046 | 0.016 | 0.005 | |
rl | 0.022 | 0.018 | 0.234 | |
d | 0.822 | 0.867 | 0.343 | |
ml × d | −0.042 | 0.012 | < 0.001 | |
g × d | –0.015 | 0.024 | 0.536 | |
rl × d | –0.124 | 0.027 | < 0.001 |
Model fit: CFI: 0.990, RMSEA: 0.017.
Estimate, Estimated unstandardized parameter value; SE, Standard error of the parameter estimate; df, Degrees of freedom; p, Probability of committing a Type I Error; γ00, Intercept of the additive predictor term; R2Marginal, Variance explained by fixed effects; R2Conditional, Variance explained by both fixed and random effects.
Significant (direct or interaction) effects (p < 0.05) are written in bold. Corresponding Bayesian models as well as models without the Linda task can be found in Appendix Tables A1, A2.