In SOC345/lehmansociology: Tools for Sociology Students Using R

knitr::opts_chunk$set(collapse = TRUE, comment = "#>", message = FALSE)

This vignette relies on the following packages and a custom dataset.

library('lehmansociology')
library(dplyr) 
library(xtable)
library(printr)
education_and_poverty<-create_educ_poverty_data()

There are many kinds of "tables" that we make in presenting statistical data and doing statistical analysis. A table can be anything that is organized into columns and rows.

For example you could consider presenting the 5 number summary for a variable to be a kind of table.

fivenum(poverty.states$PCTPOVALL_2013)

More commonly in sociology when we say table we refer to a frequency, proportion or percent table.

table(education_and_poverty$region)
# You can display this as proportions. Notice how the function builds from the inside out.
prop.table(table(education_and_poverty$region))

# Alternatively you could do this
mytable <- table(education_and_poverty$region)
prop.table(mytable)

# You can make the display show percents instead of proportions and round to one decimal.
round(prop.table(mytable)*100, 1)

# You can make it display vertically by converting the table back to a dataframe.
as.data.frame(mytable)

If we switch to a variable that is ordinal or interval we can create a cumulative percent table

cumulativetvtable <- cumsum(prop.table(table(addhealth$tvhrs)))*100
cumulativetvtable <- as.data.frame(cumulativetvtable)
round(cumulativetvtable, 0)

Now we will do acomparison (of means and medians) for different groups with a table of values.

There are many ways to do this, but this way uses some functions from the dplyr package and the magritte operator (%>%).

#start with your dataset 
education_and_poverty %>%
    #group it according the the variable you want
    group_by(region) %>% 
    # Use summarize to define the table columns
    # You can add any othe statistics you want, such as max() or min()
    # Add the new statistics to the list inside the parentheses.
    summarize(
            'Mean poverty' = mean(PCTPOVALL_2013), 
            'Median poverty' = median(PCTPOVALL_2013), 
            'IQR poverty' = IQR(PCTPOVALL_2013)
            )