arm (version 1.15-1)

rescale: Function for Standardizing by Centering and Dividing by 2 sd's

Description

This function standardizes a variable by centering and dividing by 2 sd's with exceptions for binary variables.

Usage

rescale(x, binary.inputs="center")

Value

the standardized vector

Arguments

x

a vector

binary.inputs

options for standardizing binary variables, default is center; 0/1 keeps original scale; -0.5,0.5 rescales 0 as -0.5 and 1 as 0.5; center substracts the mean; and full substracts the mean and divids by 2 sd.

Author

Andrew Gelman ; Yu-Sung Su

References

Andrew Gelman. (2008). “Scaling regression inputs by dividing by two standard deviations”. Statistics in Medicine 27: 2865--2873. https://sites.stat.columbia.edu/gelman/research/published/standardizing7.pdf

See Also

standardize

Examples

  # Set up the fake data
  n <- 100
  x <- rnorm (n, 2, 1)
  x1 <- rnorm (n)
  x1 <- (x1-mean(x1))/(2*sd(x1))   # standardization
  x2 <- rbinom (n, 1, .5)
  b0 <- 1
  b1 <- 1.5
  b2 <- 2
  y <- rbinom (n, 1, invlogit(b0+b1*x1+b2*x2))
  rescale(x, "full")
  rescale(y, "center")
  
  # For binary factor
  print(rescale(gl(2, 1), binary.inputs = "-0.5,0.5"))  # Should return c(-0.5, 0.5)

  # For a 5-point Likert scale (e.g., levels: 1, 2, 3, 4, 5)
  print(rescale(factor(c(1, 2, 3, 4, 5)), binary.inputs = "center"))  # Centered around the mean