Skip to main content
Fig. 7 | BMC Medical Research Methodology

Fig. 7

From: Comparison of analysis methods and design choices for treatment-by-period interaction in unidirectional switch designs: a simulation study

Fig. 7

Estimation biases and coverage probabilities of the four approaches (GLMM at individual level (glmm_ind), GLMM at aggregated level (glmm_agg), marginal model with GEE at individual level (gee_ind), marginal model with GEE at aggregated level (gee_agg)) using the USDo2 under the settings of \(M=48\), \(n=50\), linear effect \(\delta _j=-0.1(j-3)\), and \(\sigma _c^2 \in \{0.2, 0.5, 0.8\}\) with an exchangeable correlation structure. (vc: variance component, trt*period1-5: treatment-by-period interaction at period 1-5)

Back to article page