Shenghai Dai
  • Home
  • Research
  • R Packages
  • Projects
  • Publications
  • Teaching
  • About

R Packages

I develop R packages and research software to support reproducible quantitative and psychometric research. My software work translates methodological research into practical tools that researchers and students can use in real-world analyses.

The packages below address problems in test scoring, missing data, differential item functioning, and diagnostic classification. Together, they reflect my broader interest in connecting methodological development with accessible and reproducible research practice.

Featured Packages

subscore

Subscore computing functions in classical test theory

subscore provides functions for computing subscores and related quantities in classical test theory.

Current version: 3.3

View package →


TestDataImputation

Missing item response imputation for test and assessment data

TestDataImputation provides methods for imputing missing item responses in test and assessment data.

Current version: 2.3

View package →


DIFplus

Multilevel Mantel–Haenszel statistics for differential item functioning detection

DIFplus provides functions for detecting differential item functioning using multilevel Mantel–Haenszel statistics.

Current version: 1.1

View package →


ROCpsych

Compute and compare diagnostic test statistics across groups

ROCpsych provides functions for computing and comparing diagnostic test statistics across groups.

Current version: 1.4

View package →

Software and Reproducible Research

My software development is closely connected to my methodological research. Rather than treating software as separate from research, I use R packages to translate methodological ideas into practical tools that researchers can implement, evaluate, and reproduce.

The packages span several complementary areas:

Area Package
Subscore computation subscore
Missing data TestDataImputation
Differential item functioning DIFplus
Diagnostic accuracy ROCpsych

Together, these packages reflect a broader goal of developing accessible, reproducible, and methodologically grounded tools for quantitative research.

Related Research

These packages connect directly to my broader research in educational measurement, psychometrics, missing data, and advanced quantitative methodology.

Explore my research →

View publications →

© 2026 Shenghai Dai

 
  • WSU Large-Scale Data Lab