Alignment Linked Cell Warning Text Heading 2 Murder and* Larceny .\\ vehicle Total state* crime theft Population \9/6 Same House* 218. 722 32448...

XYZXYZ 

−6

2.4×10−6, or 0.0000024.

Model A: Assault = f(%Bachelors, %Metro, %SameHouse)

Model B: Assault = f(%Poverty, Violent, %young)

Model C: Assault = f(%Poverty, Violent, %youngmale)

Model D: Assault = f(%Poverty, Violent, %youngfemale)

Model E: Assault = f(PerCapIncome, %Unemploy, MedianAgeMale, Rain)

Model F: Assault = f(PerCapIncome, %Unemploy, MedianAgeMale, Temp)

Model G: Assault = f(PerCapIncome, %Unemploy, MedianAgeMale, Sun)

1.

a. 455.1%

b. 75.6%

c. 5.385%

d. 24.4%

a. As the percentage of state population living in a metro area increases by 1, the assault rate increases by about 209.3.

b. As the assault rate increases by 209.3, the percentage of state population living in a metro area increases by 67.3.

c. As the percentage of state population living in a metro area increases by 209.3, the assault rate increases by 1.

d. As the percentage of state population living in a metro area increases by 1, the assault rate increases by 3.11.

3.

a. Though the results for the % young variables seem consistent, it is impossible to draw any conclusions because the poverty variable is always insignificant in each model.

b. It appears that as the proportion of young males increases, the assault rate rises, and as the proportion of young females increases, the assault rate falls.

c. There seems to be an indication that the assault rate increases as there are more young people, but we cannot be confident in this result since none of the % young variables are significant at the 5% level.

d. The effect from young adults seems to relatively stable across the three models, but the results suggest that the effect from young males is larger than the effect from young females.

a. seems relatively stable (and highly significant) at about a 0.70-unit increase in the assault rate for every 1% increase in the violent crime rate.

b. cannot be assumed to be different from zero since the p-values for violent crime were so small in all models. This data seem to show that the violent crime rate has no effect on the assault rate.

c. is expected to be negative, so the only reliable model is C since it has the smallest coefficient. Models B and D, with larger coefficients, should not be used to draw inferences about violent crime rates and assault rates.

d. is unimportant since, even though it was statistically significant and of the expected sign, its coefficient estimates were always a lot smaller in magnitude than the other independent variable coefficients.

a. also positively related to the assault rate but insignificant.

b. positively related to the assault rate, but is highly significant at even the 1% level.

c. unexpectedly negatively related to the assault rate, though still insignificant at the 10% level.

d. more likely to be significantly related to the assault rate since the adjusted R-squares in these models exceed those in models B through D.

8.

9.The state of West Dakota currently has a poverty rate of 11.4%, a violent crime rate of 310.6 (per 100,000 inhabitants), and a % young male population of 15.7% (or 0.157). The state suddenly experiences a its violent crime rate, decreasing . This will likely be associated the assault rate changing about annual assaults (per , inhabitants) about ___ annual assaults ( your answer the nearest whole , decimals).