The slope of the linear regression model is:  -0.0076 
 MPG decreases by roughly -0.0076 for every additional pound.

This linear regression has a p-value of \(2.2 \times 10^{-16}\) and a Multiple \(R^2\) of 0.6926.

The Residual Standard Error in the model is 4.333 with 390 degrees of freedom.

Overall, the weight of a vehicle is a good “loose” predictor of it’s fuel-efficiency, but there are still other factors that can have a big impact. With an RSE of 4.333, the model is not an extremely accurate predictor of fuel-efficiency, yet it clearly demonstrates a statistically significant relationship of decreasing fuel-efficiency for increased weight. In this linear model, weight only predicts about 69.26% of the variance in fuel efficiency within the ISLR2 data set.

 The slope of the linear regression model is:  -0.1578 
 MPG decreases by roughly -0.1578 for every additional horsepower.

This linear regression has a p-value of \(2.2 \times 10^{-16}\) and a Multiple \(R^2\) of 0.6059.

The Residual Standard Error is 4.906 with 390 degrees of freedom.

Based on our simple linear regression, horsepower also seems like a good “loose” predictor of fuel-efficiency, with a strong, statistically significant inverse correlation to fuel-efficiency.

Overall, it seems that weight is a slightly better predictor of a vehicle’s fuel efficiency than horsepower. The simple linear regression for weight had a Residual Standard Error of 4.333, while horsepower’s was 4.906. Additionally, weight had a slightly higher Multiple \(R^2\) of 0.6926 compared to horsepower’s Multiple \(R^2\) of 0.6059.