Skip to main content
Fig. 5 | BMC Medical Research Methodology

Fig. 5

From: A graphical approach to assess the goodness-of-fit of random-effects linear models when the goal is to measure individual benefits of medical treatments in severely ill patients

Fig. 5

(Scenario 2). Benefit quantile-quantile (BQQ) plots of simulated treatment benefits at t = 4 for N = 100 patients with n = 6 measures per patient. The plots on the right panel correspond to random effects simulated from asymmetric mixtures of two bivariate normal distributions. Either bivariate component had variances \( {\sigma}_1^2={\sigma}_2^2 \) with values 1, 2, 3, 4 or 5. The left panels correspond to random effects simulated from bivariate normal distributions with the same mean and variance-covariance matrix as the corresponding non-normal distribution on the same row at the right panel

Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us