Table 2.
Regression coefficients and model fit statistics for models predicting dog populations in Tanzania. Statistics are shown for all models with δAICC < 4, ranked in order of decreasing fit as gauged by δAICC. We report the predictor variable regression coefficient estimates, model fit statistics including the partial R2FPE, the R2FPE,, adjusted R2, the degrees of freedom (df), the log-likelihood (LL), delta AIC (δAICC), and akaike weights. We also report the importance of each predictor variable (variable robustness).
| Regression Coefficient Estimates | Model Fit Statistics | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Intercept | Area (km2) | Inland vs Coastal & Island | Mainland vs. Island | Number of People | Employed as Livestock Keepers | Employed as Peasants | Partial R2FPE | R2FPE | Adjusted R2 | df | LL | δAICC | Aikakeweight |
| 5.60 | −0.708 | 0.358 | 0.448 | 57.6% | 61.6% | 72.4 | 5 | −18.65 | 0.00 | 26.0% | |||
| 9.51 | 0.124 | −0.758 | −0.447 | 0.303 | 57.9% | 63.1% | 76.3 | 6 | −17.49 | 0.94 | 16.2% | ||
| 8.07 | −0.760 | −0.219 | 0.335 | 0.300 | 57.5% | 62.7% | 76.0 | 6 | −17.65 | 1.27 | 13.8% | ||
| 9.99 | −0.907 | −0.412 | 0.281 | 55.3% | 59.5% | 75.7 | 5 | −19.39 | 1.48 | 12.4% | |||
| 11.03 | −0.767 | −0.422 | −0.489 | 0.302 | 54.7% | 60.2% | 70.3 | 6 | −18.54 | 3.05 | 5.7% | ||
| Variable robustness | 33% | 91% | 13% | 51% | 97% | 49% | |||||||