In elinw/lehmansociology: Tools for Sociology Students Using R

In this template we are going to look at the county data and create two random samples of counties, each with 200 counties. Then we will compare the samples to each other and to the combined data.

All the counties together are the population. Each sample we select will be a random sample of the population.

In statistics when we say random sample we mean each member of the population is equally likely to be selected for the sample.

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

We will create two samples of size 200, mysample1 and mysample2. Copy the example and create your own code for mysample2.

#type your code here

mysample1 <- sample_n(education_and_poverty_county, 200)

Lets look at maps of our samples.

library(maps)
county_map <- map_data("county")
state_map <- map_data("state")

# We need the FIPS values to match up the county data with the map.
data("county.fips")
county_map$polyname <- 
  paste0(county_map$region,',',county_map$subregion) 
county_map <- full_join(county_map, county.fips, by = "polyname")

In this code you have to merge the map information with the county data. The code shows you two examples, add your own code for the second sample.

county_map_all <- left_join(county_map, education_and_poverty_county, by = c("fips" = "FIPS.Code"))                 
county_map_all <-  arrange(county_map_all, group, order)                

mysample1_map <- left_join(county_map, mysample1, by = c("fips" = "FIPS.Code"))                 
mysample1_map <-  arrange(mysample1_map, group, order) 

Now let's look at where our counties are in the US. This code shows you how to get the whole US. Copy and change the code to make maps for your two samples. Make sure to add a title. Feel free to play with colors or other options.

colors<-brewer.pal(7,"Greens")
# All counties
ggplot(county_map_all,    aes(x=long, y=lat, group=group)) +
  geom_polygon(aes(fill = PCTPOVALL_2013), size = 0) +
  coord_map("polyconic" ) +
  geom_path(data = state_map, colour = "white", size = .2)+
    ggtitle( "Fig #: US Counties Poverty Rate")

How do your two samples look? Do they look randomly distributed?

One thing to know is that the mind can play tricks and does not understand the idea of random well. In a random distribution we do not expect the results to be perfectly even. We also do not expect that an individual sample will match the population perfectly.

Now let's look at some descriptive statistics for the 2 samples.

Use summary(), and var() for the two samples to compare the means of PCTPOVALL_2013 and Percent.of.adults.with.less.than.a.high.school.diploma..2009.2013.

Use table() to look at the distribution of counties by state.

Run a regression with PCTPOVALL_2013 as the dependent variable and
Percent.of.adults.with.less.than.a.high.school.diploma..2009.2013 as the independent variable.

Remember that is done with
myregression <- lm(dependent ~ independent, data = datasetname)
summary(myregression)

Remember that there is documentation here.

How do the two samples compare? What is the modal state in each? Are the other measures similar or different?

How do they compare to the complete county data?

Follow the instructions in Blackboard for sharing your data.

We are going to be working with these samples for the next few weeks. Therefore we will save them, which you do using the R function save(). Go to the Files tab and create a folder called data. Use the example code to create code to save your second sample also.

You need to run the below code, but don't knit it, which is why there is no r in the {}.

save(mysample1, file = "data/mysample1.RData")

When you are done saving you should see your two data sets in the data folder.

In the future if you want to use them you can use the load() function: load(file = "data/mysample1.RData").

Create a code chunk below that loads the two sample data sets that you saved. However you can't knit this, so run the code using the run button.

elinw/lehmansociology documentation built on May 16, 2019, 3 a.m.