How to create a new variable merging 2 dichotomous Variables?

Question

Im writing about how parents tweet about their children. My Dataset consists of a metric variable (score in emotions after sentiment analysis) and 2 dichtomous varibales (sex of the child: "son", daughter, sex of the parent: "female", "male").

Now i want to create a new categorial variable called "combination". So i get mother/daughter, mother/son, father/daughter, father/son for creating a boxplot mit means and sd (consisting of 4 groups). Does anybody have a idea how this can be done?

Answer 1

Other answers here are great. This is the lazy approach if you want an intermediate step, which will create your combination variable for use in other analyses, then try

data$combination <- paste(data$parent, data$child, sep = ".")

Answer 2

interaction is specifically meant for cases like this. It does a little bit more than paste: the result is a factor with all possible combinations of x and y as levels -- for example, if there is no "father/daughter" combination in the data, this will still be a level of the resulting factor by default (this can be turned out by setting drop = TRUE):

with(df, interaction(parent, child))

Answer 3

A quick solution would be to paste the sex labels together

set.seed(1234)
d <- data.frame(parent=c(rep('male',6),rep('female',4)), 
       child=c(rep('son',3),rep('daughter',6),rep('son',1)),
       emotion=sample(1:4,10,replace=T))

d$combination <- paste0(d$parent,'/',d$child)

Answer 4

# example dataset
df <- data.frame(
    score  = runif(100, 1, 10),
    child  = sample(c("son", "daughter"), 100, replace=TRUE),
    parent = sample(c("mother", "father"), 100, replace=TRUE)
)

# boxplot of 4 child*parent combinations
boxplot(score ~ child:parent, data=df)