Use this output to answer the questions below. Analysis of Deviance Table Model 1: Homicides ~ Police + MedHome + Mayor Model 2: Homicides ~ Police + MedHome + Mayor + Police:Mayor Resid. Df Resid....

Extracted text: Use this output to answer the questions below. Analysis of Deviance Table Model 1: Homicides ~ Police + MedHome + Mayor Model 2: Homicides ~ Police + MedHome + Mayor + Police:Mayor Resid. Df Resid. Dev Df Deviance Pr (>Chi) 1 16 406.87 15 379.47 1 27.4 0.000 f. Now we want to test if coefficient on the interactiom term between Mayor and Police is equal to 0. Using the deviance results in the table above, describe how you would construct a test statistic and what distribution that test statistic approximately follows.
]z|) (Intercept) -2.19745 0.35530 -6.185 1.31e-05 Police 0.05424 0.11390 0.476 0.640352 i MedHome -0.28076 0.05630 -4.987 0.000134 Мayor 0.60000 0.31171 1.925 0.072219 (Dispersion parameter for poisson family taken to be 1) "/>
Extracted text: The Federal Bureau of Investigation (FBI), as part of its internal security mission, serves as a database for all the reported crimes committed in the United States. One question that might be of interest is does having more police officers reduce the number of homicides? To investigate this claim, the 2013 data from the twenty largest municipal cities was supplied by the FBI's Uniform Crime Report. Below are two observations from the dataset: > head (cities) City Pop Homicides Police Density MedHome Mayor 1 New York 8175.133 335 4.2595 27016.30 5.5390 1 2 Chicago 2695.598 414 4.4309 11843.58 1.8310 The variables included in the study are: Name of city Population of city in 1,000s of people Number of homicides in the year 2013 Number of police officers per 1,000 people Number of people per square mile Median home price in $100,000 1 if city has a "Strong Mayor" form of government; 0 if not City Рор Homicides Police Density MedHome Мayor Model: log(u;) = Bo + B1Police; + B2MedHome; + B3I(Mayor;) + log(Pop;) mod1 glm (Homicides ~ Police + MedHome + Mayor, family=poisson, offset=log (Pop)) Coefficients: Estimate Std. Error z value Pr(>]z|) (Intercept) -2.19745 0.35530 -6.185 1.31e-05 Police 0.05424 0.11390 0.476 0.640352 i MedHome -0.28076 0.05630 -4.987 0.000134 Мayor 0.60000 0.31171 1.925 0.072219 (Dispersion parameter for poisson family taken to be 1)