Using the cars dataset in R, build a linear model for stopping distance as a function of speed and replicate the analysis of your textbook chapter 3 (visualization, quality evaluation of the model, and residual analysis.)
cars dataset includes 50 observations of speed and dist. distis the stopping distance in feet and speed relates to the speed of a car before applying the brakes in miles per hour.
##
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
##
## filter, lag
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union
## [1] "speed" "dist"
Plot the Stopping Distance vs Speed.
plot(cars$speed, cars$dist, xlab='Speed (mph)', ylab='Stopping Distance (ft)',
main='Stopping Distance vs. Speed')
Build a linear regression model and calculate the correlation between speed and distance.
corr = round(cor(cars$speed, cars$dist),4)
print (paste0("Correlation = ",corr))
## [1] "Correlation = 0.8069"
cars_lm <- lm(cars$dist ~ cars$speed)
cars_lm
##
## Call:
## lm(formula = cars$dist ~ cars$speed)
##
## Coefficients:
## (Intercept) cars$speed
## -17.579 3.932
plot(cars$speed, cars$dist, xlab='Speed (mph)', ylab='Stopping Distance (ft)',
main='Stopping Distance vs. Speed')
abline(cars_lm)
There is some correlation between two variables.
Summarize the linear model
summary(cars_lm)
##
## Call:
## lm(formula = cars$dist ~ cars$speed)
##
## Residuals:
## Min 1Q Median 3Q Max
## -29.069 -9.525 -2.272 9.215 43.201
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -17.5791 6.7584 -2.601 0.0123 *
## cars$speed 3.9324 0.4155 9.464 1.49e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 15.38 on 48 degrees of freedom
## Multiple R-squared: 0.6511, Adjusted R-squared: 0.6438
## F-statistic: 89.57 on 1 and 48 DF, p-value: 1.49e-12
Fitted Value vs Residuals
plot(cars_lm$fitted.values, cars_lm$residuals, xlab='Fitted Values', ylab='Residuals')
abline(0,0)
It is possible to say that the outlier values do not show the same variance of the residuals; however, it is not very clear. I think it is reasonable to continue with the analysis and assume similar variance of residuals.
qqnorm(cars_lm$residuals)
qqline(cars_lm$residuals)
The normal Q-Q plot of the residuals appears to follow the theoretical line. Residuals are reasonably normally distributed.
speed <- cars$speed
speed2 <- speed^2
dist <- cars$dist
cars_qm <- lm(dist ~ speed + speed2)
summary(cars_qm)
##
## Call:
## lm(formula = dist ~ speed + speed2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -28.720 -9.184 -3.188 4.628 45.152
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.47014 14.81716 0.167 0.868
## speed 0.91329 2.03422 0.449 0.656
## speed2 0.09996 0.06597 1.515 0.136
##
## Residual standard error: 15.18 on 47 degrees of freedom
## Multiple R-squared: 0.6673, Adjusted R-squared: 0.6532
## F-statistic: 47.14 on 2 and 47 DF, p-value: 5.852e-12
speedvalues <- seq(0, 25, 0.1)
predictedcounts <- predict(cars_qm,list(speed=speedvalues, speed2=speedvalues^2))
plot(speed, dist, pch=16, xlab='Speed (mph)', ylab='Stopping Distance (ft)')
lines(speedvalues, predictedcounts)
plot(cars_qm$fitted.values, cars_qm$residuals, xlab='Fitted Values', ylab='Residuals')
abline(0,0)
qqnorm(cars_qm$residuals)
qqline(cars_qm$residuals)
The linear model does a good job at explaining the data. Q-Q plot has some deviations, coefficients are not very significant and \(R^2\) is not increased by much. The Q-Q plot confirms that using the speed as a predictor is not sufficient to explain the data. Other factors also needs to be considered to accurately predict the stopping distance.