Modeling Seasonal and Spatiotemporal Variation: The Example of Respiratory Prescribing.

The assumptions of linear regression (normality, linearity, homoscedasticity, and absence of residual autocorrelation after adjusting for fixed and random effects) were also all adequately met ( Web Figures 3–7 ). Spatial correlation The empirical variograms of the estimated practice-level effects from the final mixed-effects model showed an increasing trend over distances up to 8 km ( Web Figure 8 ).

also in 63,971 other papers