contr.backward: Backward Difference Coding

In lbraglia/lbstat: Miscellaneous Stat-Related Utils

Backward Difference Coding

Description

mean with confidence interval and some descriptive statistics

Usage

contr.backward(k)

Arguments

k	number of levels of a factor

References

https://stats.oarc.ucla.edu/r/library/r-library-contrast-coding-systems-for-categorical-variables/

Examples

## Not run: 
contr.backward(4)
hsb2 <- read.table('https://stats.idre.ucla.edu/stat/data/hsb2.csv',
                  header = TRUE, sep = ",")

## creating the factor variable race.f
races <- c("Hispanic", "Asian", "African-Am", "Caucasian")
hsb2$race.f <- factor(hsb2$race, labels = races)

## estimates
point_est <- coefficients(summary(lm(write ~ race.f - 1, data = hsb2)))
coef_trt <- coefficients(summary(lm(write ~ race.f, data = hsb2)))
contrasts(hsb2$race.f) <- contr.backward(4)
coef_back <- coefficients(summary(lm(write ~ race.f, data = hsb2)))
diff(point_est[,1])

## End(Not run)

lbraglia/lbstat documentation built on May 9, 2024, 7:24 p.m.