Performance measures

The performance of each method was evaluated using the (absolute) bias and Mean Squared Error (MSE). The bias is calculated as the average difference between the effect estimates in the simulated samples and the true underlying effect. A negative bias indicates that the method underestimates the true underlying effect, and a positive bias indicates that the method overestimates the true underlying effect. The MSE is calculated as the average squared difference between the effect estimates in the simulated samples and the true underlying effect. The MSE represents the amount of variasbility in the effect estimates. So the higher the MSE, the higher the variability is and thus the lower the efficiency of the method is [8].