3.4 Making tables

A nice thing about numeric and factor variables is we can use the table command to see how many observations in our data fall into each category or numerical value.

## Example: how many black vs. white sounding resumes
table(resume$race)

## 
## black white 
##  2435  2435

As mentioned, factor variables have levels:

levels(resume$race)

## [1] "black" "white"

3.4.1 Crosstabulation

We can also use the table command to show a crosstabulation: a table that displays the frequency of observations across two variables.

## Example: how many black vs. white sounding resumes by call backs
## We can label the two dimensions of the table with the =
table(calledback = resume$call, race = resume$race)

##           race
## calledback black white
##          0  2278  2200
##          1   157   235

Sometimes we will want to show the proportion instead of the frequency using prop.table

## Example: proportion black vs. white sounding resumes by call backs
## Convert to proportion
prop.table(table(calledback = resume$call, race = resume$race), margin = 2) # 1 for row sum, 2 for col

##           race
## calledback      black      white
##          0 0.93552361 0.90349076
##          1 0.06447639 0.09650924

How can we interpret this crosstabulation? It should let us see the causal effect– the callback rate for each group