Statistics with R, part 2

From BioDivBorneo2010

Data types

• Independent
  • Measurements
  • Factors
• Dependent
  • Measurements
  • Counts (in different classes)
  • Species Composition

Statistical tests

A simple classification of tests, based on input data class:

independent variable:	Continuous	Nominal (Y/N, live/die..., or Categorical)
dependent variable:
Continuous (measurement)	scatterplots, cor.test, lm, glm	box plots, t.tests, wilcox.test, lm, glm
Continuous (counts)	glm(..., family=poisson)	glm(..., family=poisson), wilcox.test
Nominal or categorical	logistic regression (glm, family=binomial)	contingency tables, chisq.test, fisher.test

x <- c(1,2,3,4,4,5,6,8,9,10)
y <- c(2,3,4,4,5,6,8,8,9,11)
plot(y~x)
cor.test(x,y)
wilcox.test(x,y)
summary( lm(y ~ x) )
xy.lm <- lm(y ~ x) 
xy.lm
abline(xy.lm)

bird <- c("b","g","g","g","g","b","b","b","b","b")
build <- c("r","w","r","r","r","w","w","w","w","w")
table(bird,build)
chisq.test(bird, build)
bird2 <- c(bird,bird)
build2 <- c(build,build)
table(bird2,build2)
chisq.test(bird2, build2)