Question

3.) Now, you are going to run the multivariable linear regression model you just created.

For credit: Provide your model command and summary command below along with all the output for your model summary.

Model1 <- lm(LifeExpect2017~HouseholdIncome + Diabetic + FoodInsecure + Uninsured + DrugOverdoseMortalityRate )

> summary(Model1)

Call:

lm(formula = LifeExpect2017 ~ HouseholdIncome + Diabetic + FoodInsecure + 

    Uninsured + DrugOverdoseMortalityRate)

Residuals:

    Min      1Q  Median      3Q     Max 

-5.4550 -0.8559  0.0309  0.8038  7.1801 

Coefficients:

                            Estimate Std. Error t value Pr(>|t|)    

(Intercept)                8.266e+01  4.016e-01 205.847  < 2e-16 ***

HouseholdIncome            5.483e-05  3.365e-06  16.291  < 2e-16 ***

Diabetic                  -4.088e-01  1.683e-02 -24.292  < 2e-16 ***

FoodInsecure              -1.542e-01  1.267e-02 -12.176  < 2e-16 ***

Uninsured                 -2.242e-02  7.041e-03  -3.184  0.00148 ** 

DrugOverdoseMortalityRate -5.240e-02  3.135e-03 -16.716  < 2e-16 ***

---

Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 1.374 on 1714 degrees of freedom

  (1422 observations deleted due to missingness)

Multiple R-squared:  0.7377,    Adjusted R-squared:  0.737 

F-statistic: 964.3 on 5 and 1714 DF,  p-value: < 2.2e-16

4.) Are all your model terms statistically significant from #3? For credit, yes or no, and if no, which ones are not significant and why?

5.) For credit: Overall, is your model significantly better than nothing? If yes, explain why using the p-value approach. If no, explain why using the p-value approach.

Answer 1

4.) Are all your model terms statistically significant from #3? For credit, yes or no, and if no, which ones are not significant and why?

Ans: Yes. The all p-values are associated with the predictors are less than 0.05 i.e 5% level of significance. therefore all model term statistically significant.

5) Overall, is your model significantly better than nothing?

Ans: Yes. beacuse p value of F-statistics is p-value<2.2e-16, therefore overall model is significantly better than nothing.