I have difficulties to interpret the results of the between-country variance function.
In the random slope model, I hypothesize that the effect of internal efficacy varies across countries in a relationship with the propensity to vote among young voters. Covariance structure of the model is negative and the graph subsequently shows 'fanning in' pattern. The correlation between the random slope and intercept is negative, -.07904731. Therefore, I expect that the between-country variation will decrease by increasing internal efficacy.
However, the plot shows an increase in between-country variance.
Code: Select all
twoway function [RP2]var(cons) + 2*[RP2]cov(cons\Inteff_c)*x + [RP2]var(Inteff_c)*x^2, range(1 5) /// bgcolor(white) graphregion(color(white)) /// ytitle("Level 2 variance function") scale(0.7) /// xtitle("Internal Efficacy")
Please let me know if you have any suggestions.