Skip to contents

Getting Started with stigmaR

Source: vignettes/getting-started.qmd

Overview

stigmaR gives researchers pre-computed, state-level structural stigma scores derived from Project Implicit IAT data. The package’s three functions each merge scores into your existing dataframe by matching on state abbreviation and year.

Function What it adds
stigmaR() Pre-averaged composite indices
item_stigmaR() Individual item scores
cust_stigmaR() Your own sum-score composite built from individual items

All scores cover 50 US states + DC from 2016 onward.

Installation 

# Install from GitHub
remotes::install_github("stigmaRverse/stigmaR")

Your data requirements

Your dataframe needs one column with two-letter state abbreviations (e.g., "IL", "CA"). Everything else is up to you — participants, outcomes, covariates, etc.

The package ships a small test dataset (data/test_data.xlsx) with a state column and a person ID. We’ll use it throughout these examples:

library(readxl)
my_data <- read_excel("data/test_data.xlsx")[ , c("state", "person")]
head(my_data)
#>   state  person
#> 1    TX       1
#> 2    CA       1
#> 3    IL       1
#> 4    AL       1
#> 5    GA       1
#> 6    TN       1

stigmaR() — Composite indices

stigmaR() merges one or more pre-averaged composite indices. Each combination of year × index becomes a new column named YYYY_indexname.

my_data <- stigmaR(
  df    = my_data,
  state = "state",
  index = c("iat_sex_implicit", "iat_sex_explicit_pol"),
  year  = c("2019", "2020")
)

This adds four columns: 2019_iat_sex_implicit, 2020_iat_sex_implicit, 2019_iat_sex_explicit_pol, and 2020_iat_sex_explicit_pol.

The function prints a coverage table so you can see how many of your states matched data for each column.

Available indices (see also names(stigmaR::composite)):

  • "iat_sex_implicit" — IAT D-score (implicit pro-straight bias; higher = more stigma)
  • "iat_sex_explicit_therm" — Mean of gay men + lesbian women feeling thermometers (reversed)
  • "iat_sex_explicit_pol" — Mean of five policy opposition items
  • "iat_sex_explicit" — Omnibus mean across all explicit items

item_stigmaR() — Individual items

item_stigmaR() works identically to stigmaR() but merges raw individual items instead of composites. Use this when you want to inspect or analyze items separately before combining them.

my_data <- item_stigmaR(
  df    = my_data,
  state = "state",
  item  = c("iat_sex_imp_d", "iat_sex_exp_therm_gm", "iat_sex_exp_pol_marr"),
  year  = "2020"
)

This adds 2020_iat_sex_imp_d, 2020_iat_sex_exp_therm_gm, and 2020_iat_sex_exp_pol_marr.

A single call is capped at 100 new columns (length(item) × length(year)) to prevent accidental blowups — split larger requests into multiple calls.

All available items (see also names(stigmaR::items)):

Item name Description
iat_sex_imp_d IAT D-score
iat_sex_exp_att Explicit attitude (7-point scale)
iat_sex_exp_therm_gm Gay men feeling thermometer (reversed)
iat_sex_exp_therm_gw Lesbian women feeling thermometer (reversed)
iat_sex_exp_pol_marr Marriage rights opposition
iat_sex_exp_pol_legal Relations legality opposition
iat_sex_exp_pol_adopt Adoption opposition
iat_sex_exp_pol_serv Service refusal support
iat_sex_exp_pol_trans Transgender bathroom opposition

Respondent counts for each item are stored in parallel iat_sex_n_* columns in the items dataset (e.g., iat_sex_n_imp_d) if you want to apply your own sample-size thresholds.

cust_stigmaR() — Custom composites

cust_stigmaR() lets you define your own sum-score composite(s) from individual items. Scores are summed row-wise within each year; items cannot be mixed across years within one composite.

Single composite 

my_data <- cust_stigmaR(
  df         = my_data,
  state      = "state",
  year       = c("2019", "2020"),
  cust_index = c("iat_sex_imp_d", "iat_sex_exp_therm_gm"),
  var_name   = "implicit_plus_therm"
)
# Adds: 2019_implicit_plus_therm, 2020_implicit_plus_therm

Multiple composites at once

Pass a list to cust_index, with one name per group in var_name:

my_data <- cust_stigmaR(
  df         = my_data,
  state      = "state",
  year       = "2020",
  cust_index = list(
    c("iat_sex_imp_d", "iat_sex_exp_att"),
    c("iat_sex_exp_pol_marr", "iat_sex_exp_pol_legal",
      "iat_sex_exp_pol_adopt", "iat_sex_exp_pol_serv",
      "iat_sex_exp_pol_trans")
  ),
  var_name = c("implicit_explicit_sum", "policy_sum")
)
# Adds: 2020_implicit_explicit_sum, 2020_policy_sum

Handling missing items (na_rm)

By default (na_rm = FALSE), the composite is NA for a state-year if any component item is missing. Set na_rm = TRUE to sum whatever is available instead:

my_data <- cust_stigmaR(
  df         = my_data,
  state      = "state",
  year       = "2020",
  cust_index = c("iat_sex_imp_d", "iat_sex_exp_att"),
  var_name   = "my_sum",
  na_rm      = TRUE   # sum non-missing items; NA only if ALL items are missing
)

Output column naming

All three functions use the same naming convention: YYYY_indexname (or YYYY_itemname / YYYY_varname). This makes it easy to reference a specific year’s score while keeping multiple years in the same wide dataframe.

# Reference specific year columns directly
my_data$`2020_iat_sex_implicit`
my_data[["2020_iat_sex_explicit_pol"]]

Citation

When using stigmaR, please cite:

citation("stigmaR")

Kim, S. & Todd, N. R. (2027). stigmaR: Enhancing Accessibility, Reproducibility, and Transparency of Structural Stigma Research.

Greenwald, A. G., Nosek, B. A., & Banaji, M. R. (2003). Understanding and using the Implicit Association Test. Journal of Personality and Social Psychology, 85(2), 197–216. https://doi.org/10.1037/0022-3514.85.2.197

Further resources