
#setwd("user location") #Working directory
set.seed(123) #Standardizes the numbers generated by rnorm; see Chapter 5
n <- 100 #Number of participants; graduate students
x <- rnorm(n, 175, 7) #IV; hours since dawn
m <- 0.7*x + rnorm(N, 0, 5) #Suspected mediator; coffee consumption 
y <- 0.4*m + rnorm(N, 0, 5) #DV; wakefulness
md <- data.frame(x, m, y)
attach(md)
cor(md)

m01 <- lm(y~x, data=md)
summary(m01)
m02 <- lm(m~x, data=md)
summary(m02)
m03 <- lm(y~x+m, data=md)
summary(m03)

install.packages("ppcor")
library(ppcor)
pcor(md)
pcor.test(y,x,m)


m04 <- lm(x~m, data = md)
m04.res <- m04$residuals

m041 <- lm(y~m04.res, data=md)
summary(m041)
summary(m041)$r.squared
spcor.test(y,x,m)$estimate^2

m04.reg <- m04$fitted.values-mean(x)
m04.reg
m042 <- lm(y~m04.reg, data=md)
summary(m042)

ssreg <- sum(m04.reg^2)
sstot <- var(x)*(length(x)-1)
ssres <- sum(m04.res^2)
ssreg+ssres
sstot



m05 <- lm(m~x, data=md)
m05.res <- m05$residuals

m051 <- lm(y~m05.res, data=md)
summary(m051)

spcor.test(y,m,x)
spcor.test(y,m,x)$estimate^2 
summary(m051)$r.squared
# m 고유의 y에 미치는 영향력. 즉, m051의 r square 값

m05.reg <- m05$fitted.values-mean(m)
m052 <- lm(y~m05.reg, data=md)
summary(m052)