Skip to main content
Fig. 3 | BMC Medical Research Methodology

Fig. 3

From: Moving beyond the classic difference-in-differences model: a simulation study comparing statistical methods for estimating effectiveness of state-level policies

Fig. 3

Percent magnitude bias for all models considered in settings with small effect sizes. Note: Results showing very small grey line at 0 are equal to 0. The statistics shown will slightly favor linear over non-linear models since we have to convert magnitude bias into a total count of deaths. When magnitude bias measures are converted into the native units of the negative binomial models (log risk ratios), the negative binomial models tended to show slightly better performance relative to the linear models (as seen here)

Back to article page