Top Ranked Models and Parameters.

Note: ΔAIC is the difference in AIC value between the top ranked model and the given model. Smaller values indicate better models. Evidence Ratio can be interpreted as how much stronger the evidence is for the top ranked model over the given model. Larger values indicate stronger evidence that the top ranked model is better than the given model. β0 and β0 S.E. are estimated intercept parameter and standard error for the model (see Eqns. 5–7). β1 and β1 S.E. are estimated parameter and standard error for the deficit variable for the model (see Eqns. 5–7). α is a parameter from the negative binomial distribution that quantifies over dispersion relative to a Poisson distribution, where the over dispersion factor is defined as (1+1/α). Therefore, smaller values of α represent larger over dispersion.