In SOC345/lehmansociology: Tools for Sociology Students Using R

WebChip style means that the data have already been preprocessed to a format that is, in essence, a crosstabulation of every variable in the data set. This is very efficient when you have large amounts of individual data such PUMS data. So for example you can create a data set that contains the results of sex by race by age by income, and this will allow you to run R to recreate any specific cross tab such as sex by income. This saves a lot of space and also runs much more quickly.

Example

Using immigrationxtab first we read the dataset.

immigration <- readRDS('~/data/webchipstyle/immigrationxtab.rds')
names(immigration)

The variable called Freq represents the frequency for each of the cells. To recreate the cross tab of Region and Nativity we run

regionbynativity <- xtabs(formula= Freq ~  Region + Nativity, drop.unused.levels = TRUE, data=immigration)
regionbynativity
prop.table(regionbynativity, 1)
round(prop.table(regionbynativity, 1)*100, 0)
summary(regionbynativity)
library(MASS)
regionbynativityloglin <- loglm(Freq ~ Region + Nativity, data = immigration)
print(regionbynativityloglin)
coef(regionbynativityloglin)

Note that there are a lot of other ways to run crosstabs, but this is just to illustrate the use of WebChip style data sets that have the Freq variable.

SOC345/lehmansociology documentation built on May 9, 2019, 11:41 a.m.