Distributions, assumptions and statistical tests

H

distribution of data deviates from normality

H0

distribution of data statistically indistinguishable from normality

Test whether data normally distributed

shapiro.test(dataframe[dataframe$anothercolumnname==”rowname"]$columnname)

Example: shapiro test for

shapiro.test(penguins[penguins$species==”Gentoo”,]$body_mass)

Reject null hypothesis

p-value < 0.05

Fail to rejcet null hypothesis

p-value > 0.05

Non-parametric test

wilcox.test()

Example: Wilcox test for body mass of famle and female penguins

wilcox.test(penguins[penguins$sex==”male”,]$body_mass ,penguins[penguins$sex==”female”,]$body_mass)

Modelling relationships between columns of data

mymodel<- lm(dataframe$columnname~dataframe$anothercolumnname)

Example of modelling: flipper length and body mass of penguins

mymodel<- lm(penguins$flipper_length~penguins$body_mass)

Adding a line into a plot with the estimated coefficients of the model

abline(mymodel)