Question

only answer （e）please

2 Short Answer Solve the following problems and answer concisely. 2.1 Guns and Crime (15 points) The following regression table shows the estimation results studying the relationship between guns and crime rates in the United States (5) Violet (1) Los violent -0.000 0.05) (2) Log violent -0.2156 100 (6) Lagrobbery) -0.1276 (4) log violet) Logi t) -0.1436 0.0165) 10.12) 0.1259 0.00211 0056) 0.001) Concealed Carry law -02179 incarceration rate 0.0519 log incarceration rate 2.95 (17.68) percent male 0317 17 0.0712 income -0,0752 -35.63 Intercept 5.6678 -0,0688 DES) 4.4327 10.1456 0.4256 5.791 3911 15619 10.0723 -215.01 (106 0.3133 0.4455 0.5002 Obrations 50 The sample includes the following variables for each state in the United States where Violent: Number of violent crime incidents in a state. Robbery: Number of robbery incidents in a state Concealed carry-law: equals 1 if the state has a law that allows citizens to carry concealed weapons and to otherwise, Incarceration rate: Number of prisoners in a state (in 1.000) Percent Male: Percent of state population that is made. Income Per capita income in a state, in $1,000 (a) (2 points) Does the regression in column (1) suffer from omitted variable bias? Explain why or why not? (b) (1 points) Interpret the estimated coefficient on concealed carry-law in column (1). (e) (1 points) Interpret the coefficient on log incarceration rate in column (2) (d) (4 points) Test whether the percentage of state population that is male and the per capita income in a state are jointly significant on the 1% significance level in column (4) Define the hypothesis, test statistic, critical level and state your decision (e) (2 points) Can we compare the regression equations in columns (2) and (4)? If so, which regression provides the best fit? (1) (2 points) Can we compare the regression equations in columns (4) and (6)? If so, which regression provides the best fit? (s) (2 points) Is there statistical evidence from the regression results in column (6) that allowing citizens in a state to carry concealed weapons reduces robbery incidents? (h) (1 point)An advocacy group "For a safer America" suggests that reducing the number of prisoners in states with very high crime rates will reduce the crime rates in these states and will make America afer. Do you agree?

Answer 1

e) Yes, we can compare the regression equations in column (2) and (4), since the sample size and the dependent variable are the same. The two equations can be compared on the basis of coefficient of determination , R² .

R² tells us how much the explanatory variable is accoubtable for the variation in the dependent variable.

We will consider the equation that has a higher R² value as the best fit.

Here, the regression equation in column (4) is having a higher R² of 0.5190 than column (2) , 0.4256.

This means that 51.9 percent of the variation in the dependent variable, Log(violent) is explained by the explanatory variables by the regression equation in column (4) where as the regression equation in column (2) accounts only for 42.56 percent of the variation.

So, the regression equation of column (4) is considerd as the good fit.