Skip to contents

Gender representation audit

Source: R/audit-gender.R, R/audit-print.R
sm_audit_gender.Rd

Audits the inferred gender distribution of authors in a corpus. Gender is inferred from first names using the specified method. Results include the overall distribution, authorship-position breakdown, and a coverage metric.

This function infers a binary gender proxy from names, which has well-documented limitations. See the Limitations section in the printed output and the details below.

Usage

sm_audit_gender(
  corpus,
  method = c("genderize", "ssa", "manual"),
  api_key = Sys.getenv("GENDERIZE_API_KEY"),
  cache_dir = tools::R_user_dir("scimapR", "cache"),
  call = rlang::caller_env()
)

# S3 method for class 'sm_audit_gender'
print(x, ...)

Arguments

corpus

An sm_corpus object.

method

Character. Gender inference method:

  • "genderize": Uses the genderize.io API (requires API key for high volume).

  • "ssa": Uses US Social Security Administration frequency tables (no API needed, but US-centric).

  • "manual": Uses pre-existing inferred_gender column in the authors table (no inference performed).

api_key

Character. API key for genderize.io. Read from the GENDERIZE_API_KEY environment variable by default.

cache_dir

Character. Directory for caching API results.

call

Caller environment for error reporting.

x

An audit object to print.

...

Ignored.

Value

An sm_audit_gender S3 object containing:

distribution

Tibble with columns inferred_gender, count, pct.

by_position

Tibble breaking down gender by authorship position.

coverage

Proportion of authors with an inferred gender.

method

The method used.

confidence_summary

Summary statistics for gender confidence scores.

Details

Known limitations of automated gender inference:

  • Name-based methods assign a binary gender proxy, which does not capture the full spectrum of gender identity.

  • Accuracy varies dramatically by cultural context: names from East Asian, South Asian, and many African cultures are poorly served by Western-trained models.

  • Non-binary, transgender, and gender-diverse individuals are systematically misclassified.

  • The method cannot account for name changes, pen names, or initialised first names.

See also

Other audit: print.sm_audit_summary(), sm_audit_funding(), sm_audit_geographic(), sm_audit_oa()

Examples

corpus <- sm_example_corpus()
# Using manual method (no API call needed for examples):
gender_audit <- sm_audit_gender(corpus, method = "manual")
print(gender_audit)
#> 
#> ── <sm_audit_gender> ───────────────────────────────────────────────────────────
#> Method: manual
#> Coverage: 0% of authors have inferred gender
#> 
#> 
#> ── Overall distribution 
#> unknown: 80 (100%)
#> 
#> 
#> ── By authorship position 
#> first: NA: 233 (100%)
#> middle/last: NA: 522 (100%)
#> 
#> 
#> ── Limitations 
#> • Gender is INFERRED from first names using a binary proxy. This does not
#> capture the full spectrum of gender identity.
#> • Name-based methods have variable accuracy across cultures. East Asian, South
#> Asian, and many African names are poorly served by Western-trained models.
#> • Non-binary, transgender, and gender-diverse individuals are systematically
#> misclassified by these methods.
#> • Initialised first names (e.g., 'J. Smith') cannot be classified and reduce
#> coverage.
#> • These results should be reported with confidence intervals and
#> method-specific caveats in any publication.
#> • We recommend against using these results to make claims about individual
#> researchers' gender identity.