1. An engineer is interested in using the weight of a vehicle to predict its fue
ID: 3155086 • Letter: 1
Question
1. An engineer is interested in using the weight of a vehicle to predict its fuel efficiency. The engineer collects 79 vehicles and measures the weight (in lbs.) and the city fuel fliciey (in mps) for ach. A seatterplot of the data is given below. Bivariate Fit of City MPG By Weight 30. 25 20- 15 10-1 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 Weight The engineer fits three potential models to the data set: The engineer its three potential models to the data set Model A: A simple linear regression, yBiT Model B: A quadratic regression. y.-A' + AritA Model C: A log-log regression, log(y)BB1 log(r) ratic regression, yi = A + Ari + 2 A residual plot for each model is shown belowExplanation / Answer
a) For plot 1, there is obvious bend in scatterplot violating 'Straight enough condition'. Thus linear model is not appropriate. For plot 2, there is not much obvious bend in scatterplot, thus, linear model can be appropriate. For plot 3, there is also not much obvious bend in scatterplot. Therefore, linear model can be appropriate.
b) For plot1, residuals plot shows, obviou schanges in spread, violating 'Does the plot thicken condition'. Therefore variance is not constant. For plot 2, 'Does the plot thicken condition' is not violated, therefore, variance is constant and for plot 3, variance is constant.
c) Model 2 is most appropriate.
Related Questions
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.