Table 5 Estimated regression coefficients and the standard errors (in parentheses) for the Poisson, NB, ZIP and ZINB models in the real data application
From: A comparison of residual diagnosis tools for diagnosing regression models for count data
Variables | Poisson | NB | ZIP | ZINB |
|---|---|---|---|---|
Black vs. others | 0.188(0.085)* | − | 0.300(0.097)* | − |
Chronic conditions | 0.221(0.020)** | 0.217(0.026)** | 0.216(0.023)** | 0.217(0.027)** |
Self-perceived health | ||||
Excellent vs. poor | -1.093(0.190)** | -1.089(0.216)** | -1.028(0.207)** | -1.089(0.216)** |
Average vs. poor | -0.505(0.074)** | -0.478(0.100)** | -0.451(0.088)** | -0.478(0.101)** |
Limited activities | 0.453(0.070)** | 0.464(0.087)** | 0.426(0.077)** | 0.464(0.087)** |
Years of education | -0.017(0.008)* | -0.023(0.010)* | -0.019(0.009)* | -0.023(0.100)* |