I'm working in R-studio a linear regression of a data sample of 30. The relationship is between the weight of a car and its horsepower. I've done almost all of the work but I'm stuck in two questions. Let me know if the first one is right or if not what I'm missing.
- Show the appropriate graphs to evaluate the assumptions of homogeneity, normality of errors, and uncorrelated errors. What do the plots suggest?
Residuals vs Fitted: The residuals are randomly scattered around the line. So it can be said that the relationship is linearly reasonable. second.
Normal Q-Q: The residuals appear to be normally distributed since most of the points form a straight line.
Scale-location: The residuals appear to be randomly scattered around a line that is approximately horizontal, indicating that the errors have constant variance.
Residuals vs leverage: There are no influential cases because there are no points outside the dashed line.
- When the weight of a car increases by 100 pounds, what happens to the power of the car? Perform an analysis using the fitted regression line.
Any suggestion of how to work on this in R-studio?