Lab Assignment 5

setwd("C:/Users/Qiu J/Desktop/MSSP+DA 2021FALL/MSSP 897-002 Applied Linear Modeling/Assignment/Lab Assignment 5")
NLSY <- read.csv("C:/Users/Qiu J/Desktop/MSSP+DA 2021FALL/MSSP 897-002 Applied Linear Modeling/Assignment/Lab Assignment 5/NLSY-3.csv")

Using the added variable plot method, assess whether the highest grade completed by the child’s mother (medu) is an omitted relevant variable in the following regression model: read = magebirth + breastfed (include all relevant plots in your answer)

lm1 <- lm(read ~ magebirth + breastfed, data=NLSY)
resid_read <- lm1$residuals

lm2 <- lm(medu ~ magebirth + breastfed, data=NLSY)
resid_medu <- lm2$residuals

plot(density(resid(lm1)))

plot(density(resid(lm2)))

qqnorm(resid(lm1))
qqline(resid(lm1))

qqnorm(resid(lm2))
qqline(resid(lm2))

The range and distribution of the two regressions are different. Residuals of the initial model conform to the normal distribution, but residuals of the second regression have two peaks.

plot(lm1$residuals,lm2$residuals)
abline(lm(lm2$residuals~lm1$residuals),col="red")

lines(lowess(lm2$residuals~lm1$residuals),col="blue")

The regression and lowess line are not horizontal. There is reason to suspect that medu is an omitted variable.