chapter: 4 相関の効果量

後で加筆。

4.0.1 Pearson’s r

# データ生成
library(mvtnorm)
set.seed(123)
sigma <- matrix(c(100,50,50,100), byrow=TRUE, ncol=2) 
mu <- c(50, 60)
n <- 100
dat <- data.frame(rmvnorm(n=n, mean=mu, sigma=sigma)) 
colnames(dat) <- c("A","B")

# ピアソンの積率相関
library(correlation)
correlation(dat)

## # Correlation Matrix (pearson-method)
## 
## Parameter1 | Parameter2 |    r |       95% CI | t(98) |         p
## -----------------------------------------------------------------
## A          |          B | 0.36 | [0.18, 0.52] |  3.87 | < .001***
## 
## p-value adjustment method: Holm (1979)
## Observations: 100

4.0.2 ベイズ

correlation(dat, bayesian = T)

## # Correlation Matrix (pearson-method)
## 
## Parameter1 | Parameter2 |  rho |       95% CI |      pd | % in ROPE |         Prior |        BF
## -----------------------------------------------------------------------------------------------
## A          |          B | 0.34 | [0.18, 0.51] | 100%*** |     0.50% | Beta (3 +- 3) | 162.40***
## 
## Observations: 100