R Observatory: Metrics for aiDIF version 0.3.0

Balamuta, James Joseph

doi:10.5281/zenodo.21843040

aiDIF

0.3.0Package record

Differential Item Functioning for AI-Scored Assessments

0packages dependsuggested by 1

2.5Kdownloads / year

80.6%test coverage

13/13flavors OK

Overview

About

Maintained by Subir HaitCRAN page ↗GitHub ↗Issues ↗

Detects and quantifies differential item functioning (DIF) in AI-scored educational and psychological assessments. Provides a fully self-contained robust DIF engine (M-estimation via iteratively re-weighted least squares with the bi-square loss), cross-checked against the reference implementation in 'robustDIF', alongside the Differential AI Scoring Bias (DASB) test, which detects item-level scoring shifts that differ across subgroups when comparing human and AI scoring conditions. Supports independent and paired scoring designs, including a bootstrap estimator of the cross-condition covariance that paired designs require, a one-call workflow from raw item scores through 'mirt' calibration, robust linking of the cross-condition contrast, bootstrap standard errors, multiplicity control, simulation of paired human and AI item scores, anchor weight diagnostics, and an AI-effect classification framework. Methods follow Halpin (2024) doi:10.1007/s11336-024-09957-6.

Changes in 0.3.0

New features

fit_aidif_responses() runs the whole analysis from raw item scores: it calibrates both scoring conditions in 'mirt' (item parameters free in both groups), converts the fits with as_aidif(), estimates the cross-condition covariance for paired designs, and calls fit_aidif(). Dichotomous (2PL) and polytomous (graded) items are supported.
paired_cross_cov() estimates the cross-condition covariance that design = "paired" requires, by resampling examinees within group and refitting both calibrations on each resample. In 0.2.0 the paired design accepted this matrix but offered no way to obtain it, so it was usable only by analysts who could derive it themselves. The default estimator rescales the bootstrap correlation by the model-based standard errors, which guarantees a non-negative paired variance. Bootstrap replicates can run in parallel on all platforms via cores.

Show all changes in 0.3.0Hide the rest of the changes in 0.3.0

fit_aidif_responses() defaults to metric = "linked". When the AI engine changes item slopes and the groups differ in mean trait, raw DASB is shifted on every item; in simulation (AI slopes 0.8 times human, impact 0.5 SD, no true DASB) the raw test rejected about 16% of the time at a nominal 5%, while the linked metric held 5%. fit_aidif() keeps its 0.2.0 default ("common") for backward compatibility.
fit_aidif_responses(se = "bootstrap") reports bootstrap standard errors for DASB. Under metric = "linked" the item-common offset is re-estimated in each replicate, so these standard errors include the variability of the bi-square weights that the analytic linked-metric variance ignores.
simulate_aidif_responses() generates paired human and AI item scores from a logistic graded response model with known DIF, DASB, AI drift, AI slope change, and human-AI agreement (Gaussian copula). Unlike simulate_aidif_data(), which draws parameter estimates directly, it produces response data that go through real calibration, so it supports Type I error, power, and coverage studies of the full workflow.
compare_robustdif() runs aiDIF's robust scaling engine and 'robustDIF' on the same estimates for each scaling convention and reports the difference. The two agree to machine precision.

Bug fixes

print() for DASB tables dropped the item labels and printed row numbers instead. Fixed.

Other changes

'robustDIF' added to Suggests; 'parallel' added to Imports.
Package URL moved to <https://github.com/subirhait/aiDIF>.

From NEWS.md in the release

Install

R (CRAN)pak (dev)Ubuntu (r2u)Fedora (COPR)openSUSE (OBS)

install.packages("aiDIF", repos = "https://cloud.r-project.org/")

Installs the current CRAN release. Prebuilt binaries on Windows and macOS.

if (!requireNamespace("pak", quietly = TRUE))
  install.packages("pak", repos = "https://cloud.r-project.org/")
pak::pak("subirhait/aiDIF")

Development version from subirhait/aiDIF. Installs pak first if it is not already available.

Release

# enable cranapt (r2u) if it is not already set up, then install
grep -rqs r2u.stat.illinois.edu /etc/apt/sources.list /etc/apt/sources.list.d/ \
  || curl -fsSL https://raw.githubusercontent.com/eddelbuettel/r2u/master/inst/scripts/add_cranapt_noble.sh | sudo bash
sudo apt install -y r-cran-aidif

Fast Ubuntu binaries via r2u (no compilation). The cranapt setup runs once and is skipped when it is already enabled. We list this channel from recorded r2u download activity rather than from a live index of the repository, so treat it as last known rather than guaranteed.

bspm (used by r2u) also needs make: sudo apt install make

# enable the iucar/cran COPR if it is not already present, then install
ls /etc/yum.repos.d/*iucar*cran*.repo >/dev/null 2>&1 \
  || sudo dnf copr enable -y iucar/cran
sudo dnf install -y R-CRAN-aiDIF

RPMs from the iucar/cran COPR, rebuilt from CRAN.

For install.packages() integration in R, also: sudo dnf install R-CoprManager

# add the autoCRAN OBS repo (Tumbleweed) if it is missing, then install
zypper -q lr | grep -qi autoCRAN \
  || sudo zypper addrepo https://download.opensuse.org/repositories/devel:/languages:/R:/autoCRAN/openSUSE_Tumbleweed/devel:languages:R:autoCRAN.repo
sudo zypper install -y R-aiDIF

RPMs auto-built from CRAN on openSUSE OBS (Tumbleweed). We list this channel from recorded autoCRAN download activity rather than from a live index of the repository, so treat it as last known rather than guaranteed.

Health

CRAN checks

13OK

CRAN check page ↗

Slowest check: 4.0 min · r-oldrel-windows-x86_64

Status changes

OK13 OK2026-08-05
NOTE12 OK, 1 NOTE2026-08-01
OK8 OK2026-04-22

Advisories

none known

How this is measured·How advisories are found

Documentation

In release 0.3.0

READMEYes · 465 wordsVignettesYes · 1 · evaluation depends on the environmentNEWSYes · 100% structured · mentions 0.3.0Citation fileYes · computed when R runs itWebsitenot declared in DESCRIPTION

Help pages: 30 (13 internal)
Examples that run: 85%11 run in a plain check (2 of them under a condition) · 2 only when donttest examples are enabled (R CMD check --as-cran)
Exported functions with examples: 92%
Documented parameters: 93%
Return-value docs: 100%
Help pages citing references: 7%

Vignettes (1)

Introduction to aiDIF HTML

How these are measured

Downloads

2.5K

CRAN downloads in the past year

Rank #8,753 · ~7/day · ~206/mo

Releases R versions

Daily download trend is not available in this view yet.

39430 days

1.1K90 days

2.5K1 year

Compare downloads with other packages →

Also on59 r2u26 autocran

How downloads are counted

Repository

subirhait/aidif ↗Account: subirhait

0Stars

0Forks

0Open issues

0Open PRs

0Releases

Last activity 2026-10-21

Responsiveness

not measured

Earlier repositories

github.com/causalfragility-lab/aidif · last recorded on 2026-10-22

How these are measured

PracticesBeta

Repository github.com/subirhait/aidif is linked, but repository-practices checks have not run for it yet.

How this is detected·Detection ruleset v4 (2026-10-20)

Dependencies

Declared dependencies

5 external dependencies (excludes base and recommended)

Installing aiDIF adds no packages beyond those that come with R.

Depends (1)

R >= 3.5.0

Imports (5)

Matrixstatsgraphicsutilsparallel

LinkingTo (0)

none

Suggests (5)

mirt robustDIF testthat knitr rmarkdown

Enhances (0)

none

Dependency license breakdown →

Reverse dependencies

0direct

0indirect

Suggested by 1

Suggests, most downloaded first

aiEvalR

How these are counted

Code & Tests

People & History

People (1)

Subir Hait

Author, Maintainer since 0.1.0 (2026)

Releases (3)

Pick two releases to compare their code metrics.

0.3.0Latest
record
2026-10-22 · current release · diff ↗ ·
added 4 exports
compare_robustdif, fit_aidif_responses, paired_cross_cov, simulate_aidif_responses
·
26 hours in CRAN's queue
Listed in CRAN's incoming queue from 2026-10-21 19:49 to 2026-10-22 21:24 UTC, in 7 readings.
0.2.0
record
2026-09-05 · diff ↗ ·
added 1 export
as_aidif
R
R 4.6.0 released · 2026-04-24
0.1.0
record
2026-04-21
R
R 4.5.0 released · 2025-04-11

Across CRAN, the median accepted release of the past year spent 26 hours in the queue, 90% under 11 days.How queue time is measured

Package metadata

First published: 2026-04-21
Total releases: 3 / 1 yrs
License: GPL (>= 3) OSI
Minimum R: ≥ 3.5.0
Download size: 102 KB

Cite

Cite this package

Run in R for the authors' preferred citation:

citation("aiDIF")

The package provides a citation file that builds its entry when R runs it. Run citation() to see it.

Cite the R Observatory

For a number measured here: a download total, a coverage figure, an archival date.

Format

APA

Balamuta, J. J. (2026). R Observatory: Metrics for aiDIF version 0.3.0 [Data set]. HJJB, LLC. Data release v2026-10-24. https://doi.org/10.5281/zenodo.21843040

From data release v2026-10-24, which the citation names so these numbers can be found later. More on citing and the projects behind them.

Report a problem with this page →