Question: Multiple linear regression Help using R creating a linear regression to model how the number of calories in one portion of cereal depends on various
Multiple linear regression
Help using R creating a linear regression to model how the number of calories in one portion of cereal depends on various variables. The data set for the regression isUScerealfrom the packageMASS. To access the data set, load the packageMASSand simply typeUScerealin the console.
you can get the data from:
> library(MASS)
> data = UScereal
- Estimate the coefficients of the following model:calories=0+1protein+2fat+3fibre+4carbo+5sugars+.
- The variablefibreis not statistically significant. Does that make sense?
- Using thesummaryoutput as well as the diagnostic plots, explain what indicates that this is a reasonable model (give at least three reasons why this is a reasonable model).
- Using thesummaryoutput as well as the diagnostic plots, explain what potential problems with this model are and how they might affect the model (list at least three potential problems and how they might affect the model).
- Identify three largest outliers (in absolute value) of this model.
- Remove the three largest outliers (in absolute value) of this model and run the same multiple linear regression. Using thesummaryoutput as well as the diagnostic plots, explain how the removal of the three largest outliers improves the model (list at least four improvements)
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
