Data Preparation

Load Necessary Packages

library(tidycensus)
library(tidyverse)
library(scales)
library(plotly)
library(mapview)

Retrieve ACS Data

First, I will use the tidycensus function get_acs() to locate data based on parameters for spatial resolution, table, year, and state.

degrees_il <- get_acs(
  geography = "county",
  
  # give the variable a more human-readable name
  variables = c(percent_graduate = "DP02_0066P"),
  year = 2023,
  state = "IL"
)

# arrange percent estimates in descending order
arrange(degrees_il, desc(estimate))

## # A tibble: 102 × 5
##    GEOID NAME                       variable         estimate   moe
##    <chr> <chr>                      <chr>               <dbl> <dbl>
##  1 17019 Champaign County, Illinois percent_graduate     25.5   1.1
##  2 17043 DuPage County, Illinois    percent_graduate     21     0.4
##  3 17097 Lake County, Illinois      percent_graduate     19.5   0.5
##  4 17031 Cook County, Illinois      percent_graduate     17.6   0.2
##  5 17077 Jackson County, Illinois   percent_graduate     16.8   1.2
##  6 17113 McLean County, Illinois    percent_graduate     15.7   0.8
##  7 17109 McDonough County, Illinois percent_graduate     15.1   1.7
##  8 17167 Sangamon County, Illinois  percent_graduate     14.2   0.6
##  9 17133 Monroe County, Illinois    percent_graduate     14     1.9
## 10 17143 Peoria County, Illinois    percent_graduate     13.7   0.8
## # ℹ 92 more rows

Champaign County has the largest estimated percentage of graduate degree holders at 25.5% and a margin of error of 1.1%. DuPage and Lake counties follow with margins of error at 0.4% and 0.5% respectively. Rows 1-10 are currently visible out of 102. It’s difficult to conceptualize the differences between counties by viewing the table, so I will next create a plot depicting all 102 counties with their estimates.

Analysis Part A: Visualization & Analysis with ggplot and plotly

To make the data for all 102 counties visible, I’ll plot the data with ggplot().

# since Illinois has 102 counties, in order to not have y-axis labels values overlap, the dimensions for this code chunk's output are specified as:
# #| fig-height: 15 
# #| fig-width: 10
# #| dpi: 300

degree_plot_errorbar <- ggplot(degrees_il, aes(x = estimate,
                                      y = reorder(NAME, estimate))) +
  
  # create and stylize the error bars
  geom_errorbar(aes(xmin = estimate - moe, xmax = estimate + moe),
                width = 0.5, linewidth = 0.5, color = "black") +
  
  # similarly, create points
  geom_point(color = "blue", size = 1.5) +
  scale_x_continuous(labels = label_percent(scale = 1)) +
  
  # clean y-axis labels
  scale_y_discrete(labels = function(x) str_remove(x, " County, Illinois")) +
  
  # customize labels
  labs(
    title = "Percent of Population with Graduate Degrees, 2023 ACS",
    subtitle = "Counties in Illinois",
    caption = "Data acquired with R and tidycensus. Error bars represent margin of error around estimates.",
    x = "ACS Estimate",
    y = "") +
  
  theme_minimal()

degree_plot_errorbar

With the values all plotted, it becomes clear that while Champaign County ranks in at first place for estimated percent of population with a graduate degree, Brown County comes in last place, followed by Perry and Shelby counties at second- and third-to-last, respectively.

We can see that Brown County’s percent value is somewhere between 3% and 4%, but we can’t tell exactly what it is just by looking at this plot.

To provide the exact percent values for each point, I will plot the data in an interactive plot using the ggplotly() function.

ggplotly(degree_plot_errorbar, tooltip = "x", height = 1700)

Now we can see that Brown county’s percent estimate is 3.4% when hovering over the data point. However, its margin of error bar shows that the true percent value could be anywhere from around 2% to 4%, which means it’s possible that Brown County could actually place anywhere in the bottom 30-40 counties, as these counties all show overlapping margins of error with Brown County (with the exception of Marion County).

Analysis Part B: Spatial Visualization & Analysis of Income Data

Now that I have created an interactive plot and analyzed the data for the percent of graduate degrees in Illinois Counties, I will explore a variable from the ACS data that often correlates with education status: income.