***Zero inflated poisson Model without a breakpoint:***
Call:
zeroinfl(formula = V1 ~ daynum, data = awb.lar)
Pearson residuals:
Min 1Q Median 3Q Max
-0.3672 -0.3230 -0.2850 -0.2482 13.5661
Count model coefficients (poisson with log link):
Estimate Std. Error z value Pr(>|z|)
(Intercept) -0.3233952 0.1691332 -1.912 0.0559 .
daynum -0.0003241 0.0005245 -0.618 0.5366
Zero-inflation model coefficients (binomial with logit link):
Estimate Std. Error z value Pr(>|z|)
(Intercept) 1.274648 0.204328 6.238 4.43e-10 ***
daynum -0.001549 0.000674 -2.298 0.0216 *
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Number of iterations in BFGS optimization: 15
Log-likelihood: -336.2 on 4 Df
---
***Zero inflated poisson Model with a breakpoint corresponding to the expiration of the assault weapon ban:***
Call:
zeroinfl(formula = V1 ~ awb/daynum, data = awb.lar)
Pearson residuals:
Min 1Q Median 3Q Max
-0.5585 -0.3072 -0.2559 -0.2117 12.4700
Count model coefficients (poisson with log link):
Estimate Std. Error z value Pr(>|z|)
(Intercept) 0.1335902 0.5253115 0.254 0.799
awbTRUE -0.7648132 0.8413005 -0.909 0.363
awbFALSE:daynum 0.0008780 0.0016418 0.535 0.593
awbTRUE:daynum 0.0007345 0.0032309 0.227 0.820
Zero-inflation model coefficients (binomial with logit link):
Estimate Std. Error z value Pr(>|z|)
(Intercept) 2.702555 0.530950 5.090 3.58e-07 ***
awbTRUE -0.602907 0.963816 -0.626 0.5316
awbFALSE:daynum 0.002767 0.001712 1.616 0.1061
awbTRUE:daynum -0.009344 0.004572 -2.044 0.0410 *
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Number of iterations in BFGS optimization: 18
Log-likelihood: -325.7 on 8 Df
---
----
***Comparing zero inflated poisson model to normal poisson without a breakpoint:***
Vuong Non-Nested Hypothesis Test-Statistic: -2.528253
(test-statistic is asymptotically distributed N(0,1) under the
null that the models are indistinguishible)
in this case:
model2 > model1, with p-value 0.005731591
---
***Comparing zero inflated poisson model to normal poisson with a breakpoint:***
Vuong Non-Nested Hypothesis Test-Statistic: -2.521703
(test-statistic is asymptotically distributed N(0,1) under the
null that the models are indistinguishible)
in this case:
model2 > model1, with p-value 0.005839407
---
---
***Likelihood ratio test comparing the zero inflated models with and
without breakpoint***Likelihood ratio test
Model 1: V1 ~ daynum
Model 2: V1 ~ awb/daynum
#Df LogLik Df Chisq Pr(>Chisq)
1 4 -336.21
2 8 -325.69 4 21.032 0.0003121 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Vuong Non-Nested Hypothesis Test-Statistic: -1.54627
(test-statistic is asymptotically distributed N(0,1) under the
null that the models are indistinguishible)
in this case:
model2 > model1, with p-value 0.0610197
Vuong Non-Nested Hypothesis Test-Statistic: -1.772001
(test-statistic is asymptotically distributed N(0,1) under the
null that the models are indistinguishible)
in this case:
model2 > model1, with p-value 0.03819723
......
Rrrr! Avast Ye, YO HO!
