Mapping Demographic Disparity Results

[tiernanmartin]

I would like to be able to display racial demographic disparity scores in a choropleth map, but the geo_bias_data object does not include these variables. Can this information be made available to the user?

[Deckart2]

Hi @tiernanmartin! Apologies for the slow response and thank you for the message!

You are absolutely correct. The geo_bias_data by default does not output any racial/ethnic demographic results. However, if you are interested in getting geographic disparity scores using a population not included in the tool, the tool can handle this functionality by taking a supplemental geographic dataset. The description/example code in our documentation shows code syntax from the sedtR::call_sedt_api() function.

To expand a bit on that information, I suspect your workflow should be something like:

1. Identify and download point resource data and determine the appropriate geographic scale (sounds like you have already done this)
2. Get racial/ethnic population count data for the geography of interest. I recommend tidycensus.
3. Use the sedtR::call_sedt_api() function and specify optional arguments geographic_file_path, geographic_geo_id_column and geographic_columns() referencing the data you pulled in step 2.
4. View your results in the geo_bias_data object

This workflow works for me:

library(sedtR)
library(tidyverse)
library(tidycensus)

#Step 1a: Identify point resource data
download.file(
  "https://equity-tool-api.urban.org/sample-data/minneapolis_bikes.csv",
  destfile = "bikes.csv")

#Step 1b: determine appropriate geographic scale:
# Scale is the CITY because Minneapolis is a city 

# Step 2: Get racial/ethnic population. In this case, we will do non-Hispanic Black population
minneapolis_black_pop <- tidycensus::get_acs(
  geography = "tract",
  variables = "B02001_003",
  state = "MN",
  place = "Minneapolis",
  year = 2022,
  geometry = FALSE) |>
  rename(black_pop_estimate = estimate,
         black_pop_moe = moe)

write_csv(minneapolis_black_pop, "minneapolis_black_pop.csv")

#Step 3: Call API with geographic supplemental data arguments
sedt_results <- sedtR::call_sedt_api(
  resource_file_path = "bikes.csv",
  resource_lat_column = "lat",
  resource_lon_column = "lon",
  geographic_file_path = "minneapolis_black_pop.csv",
  geographic_geo_id_column = "GEOID",
  geographic_columns =  list(
    black_pop_estimate = "black_pop_moe")
)

geo_bias <- sedt_results$geo_bias_data

#Step 4: Visualize geographic disparity scores
map <- sedtR::create_map(sedt_results$geo_bias_data, 
           interactive = FALSE,
           save_map = FALSE, 
           col_to_plot = "diff_black_pop_estimate_geographic")

[tiernanmartin]

Perfect! I can't believe I didn't think to do that. Thanks for going the extra mile by providing a reproducible example.