In lehmansociology/lehmansociology: Tools for Sociology Students Using R

library(ggplot2)
library(lehmansociology)
library(grid)
library(scales)
library(magrittr)
library("dplyr")
library(googlesheets)
library(broom)
library(xtable)
library(gridExtra)
# Set options for nicer looking documents
options(xtable.comment = FALSE)
knitr::opts_chunk$set(message=FALSE, warning=FALSE)

replaceCommas<-function(x){
  x<-as.numeric(gsub("\\,", "", x))
}

In this exercise we are going to explore how region of the country relates to the relationship of education and income.

First we create the data set.

# Set up some data
poverty13 <- select (poverty.states, FIPStxt, Area_Name, PCTPOVALL_2013, PCTPOV05_2013,
                     MEDHHINC_2013, Rural_urban_Continuum_Code_2013)
poverty13$FIPS.Code <- as.integer(poverty13$FIPStxt)
poverty13$MEDHHINC_2013 <- replaceCommas(poverty13$MEDHHINC_2013)

lessthanhighschool13 <- select(education.states, Area.name, FIPS.Code,
                            Percent.of.adults.with.less.than.a.high.school.diploma..2009.2013,
                            Percent.of.adults.with.a.bachelor.s.degree.or.higher..2009.2013,
                            Percent.of.adults.with.less.than.a.high.school.diploma..2000,
                            Percent.of.adults.with.a.bachelor.s.degree.or.higher..2000
                          )

education_and_poverty <- merge(poverty13, lessthanhighschool13, by.x='FIPS.Code', by.y='FIPS.Code')

#type your code here
# First let's create the region data set
gs_region<-gs_url('https://docs.google.com/spreadsheets/d/1h_jY4A44WoSLkrqhwZZ9oJh51N2GybwVvGgEaY3n2gc/pubhtml')
region_data<-gs_read(gs_region)
# We need to change this column name because the map data uses the term region differently.

# Add the region variable to education_and_poverty by matching the FIPS code
education_and_poverty <- merge(education_and_poverty, region_data, 
                               by.x='FIPS.Code', by.y='FIPS.Code')

Deal with commas in Median Household Income

replaceCommas(education_and_poverty$MEDHHINC_2013)

This scatterplot shows the relationship between the percent of the population with less than a high school diploma and the overall poverty rate. Notice that the color of the points indicates the region. This is one way to think in a more multivariate (more than two at a time) manner.

Make sure to add an appropriate title and labels.

baseplot <- ggplot(education_and_poverty,
       aes(x = Percent.of.adults.with.less.than.a.high.school.diploma..2009.2013,
           y = PCTPOVALL_2013,
           fill = region
           )
       ) +
  geom_point(aes(color = region)) +
  ggtitle("") +
  labs(x = "",
       y="")
baseplot

Summarize here what the graph tells you about the three variables and how they relate to each other.

Previously we had learned how to make a table showing univariate statistics by region.

First we group the data by region.

databyregion <- education_and_poverty  %>% 
  group_by(region)

# You can add any othe statistics you want, such as max() or min()
# Add the new statistics to the list inside the parentheses.
comparison_poverty <- databyregion %>% 
  summarize(Mean = mean(PCTPOVALL_2013), Median = median(PCTPOVALL_2013), IQR = IQR(PCTPOVALL_2013), Variance = var(PCTPOVALL_2013))

 comparison_poverty

Now compare the regression results for the different regions. To do this we us our grouped data and some advanced code that uses "do" to loop through the regions.

regbyregion<-databyregion  %>% 
  do(model = lm(PCTPOVALL_2013 ~ Percent.of.adults.with.less.than.a.high.school.diploma..2009.2013, data = .))

results <- tidy(regbyregion, model)
results

rsquared <- glance(regbyregion, model)
rsquared

Nicer formatting when you knit, comment out the other ones if you want to use this, otherwise comment out this.

print(xtable(results))
print(xtable(rsquared))

Let's make separate plots by region. Feel free to add a regression line if you wish.

# Since we have 4 regions, let's use the the grid.arrange function from gridExtra to set up 2 columns and 2 rows.

plotbyregion <- ggplot(education_and_poverty,
       aes(x = Percent.of.adults.with.less.than.a.high.school.diploma..2009.2013,
           y = PCTPOVALL_2013,
           fill = region
           )
       ) +
  geom_point(aes(color = region)) +
  # Notice that we are adding this so that the graphs are broken out by region
  facet_grid(. ~ region ) +
  ggtitle("") +
  labs(x = "",
       y="")
plotbyregion

Based on the scatter plots how would you describe the similarities and differences of the regions?

Examing the intercepts and coefficients. How similar or different are they? Remember that the intercept represents the predicted value of Y (Y hat) when X is equal to 0. How do the intercepts of the regions compare to each other?

Are all of the coefficients for Percent with less than a high school diploma the same sign?

If not, which are positive and which are negative?

How do the values for R squared compare to each other?

Summarize what the data show about the relationship between the variables by region.

Now do the same analysis with a different pair of independent and dependent variables.

lehmansociology/lehmansociology documentation built on May 21, 2022, 9:06 p.m.